"Vibe-Trading: Your Personal Trading Agent"
Python ai-agentalgorithmic-tradingbacktestingfintechllmmcpmulti-agentpythonquantitative-financetrading
release: 0.1.16 — a number that can show where it came from
Haozhe Wue1dbea8ad307
diff --git a/CHANGELOG.md b/CHANGELOG.mdindex 4241d6aaf..6abd0a394 100644--- a/CHANGELOG.md+++ b/CHANGELOG.md@@ -5,29 +5,52 @@ This project adheres to [Semantic Versioning](https://semver.org/).## [Unreleased]+## [0.1.16] — 2026-09-29++Rolls up 492 commits / 116 merged pull requests since 0.1.15, from 16+contributors.++The theme of this cycle is a number that can show where it came from. The+final-answer grounding gate stopped guessing what a number is from the words+around it: the model declares each figure's role — observed, derived,+proposed, cited or count — the gate checks that declaration against the+session's tool evidence, and a figure that fails is cut from the answer+instead of the whole answer being refused. A backtest's own output (its+Sortino, turnover, weights and Monte Carlo p-values, not only its Sharpe)+now grounds the report written about it, run cards cite a metric only where+the saved CSV agrees, and the generated broker matrix states each profile's+permissions as declarations rather than as runtime verification. Research+runs stopped losing what they fetched: compaction follows the model's real+context window instead of a fixed 40K estimate that began clearing tool+results from the third step. The missing-data sweep reached 59 more alphas:+15 qlib158 count factors, 20 that filled a missing input with a constant+and 24 comparison gates. Argentina (BYMA), weekly and monthly bars, Bahasa+Indonesia, Gildata, OpenCode, Email and WebSocket channels and four broker+connectors (KIS, Upbit, Toss Securities, Scalable Capital) land alongside.+### Added-- **Stock profiles carry the listing's own identity** (#1577). `get_stock_profile`- returns a `listing` block beside the issuer's fundamentals: Yahoo's symbol,- exchange, quote type and quote currency, plus the financial currency and- underlying symbol when Yahoo declares them. A secondary listing no longer- reads as the primary one. `GOOGL.BA` is an ARS line in Buenos Aires carrying- Alphabet's USD financials, and `GGALD.BA` is a USD line whose issuer reports- in ARS. Nothing is inferred from the ticker.--- **Volume statistics in `technical_indicators`** (#1571). The tool returns the- latest volume, its 20-bar mean and their ratio, computed from the bars it- already fetched. The mean needs a complete 20-bar window. `volume.unit` is- the unit the serving source declares (board lots on the A-share sources,- shares on the Yahoo family, `null` when undeclared), because the two are- 100x apart and nothing in the numbers tells them apart.--- **Argentina (BYMA) market data** (#1543). A `.BA` symbol — a BYMA listing or a- locally traded CEDEAR — is its own market `ar_equity`, quoted in ARS, served- by `yahoo` → `yfinance` → `local`, and reported as market `ar` by- `get_stock_profile`. Execution is deliberately not modelled: an Argentine- backtest raises instead of borrowing US or crypto commissions, lot sizes,- settlement and short-selling rules. The Web UI carries the market too —+- **Stock profiles carry the listing's own identity** (#1577, thanks @Shizoqua).+ `get_stock_profile` returns a `listing` block beside the issuer's+ fundamentals: Yahoo's symbol, exchange, quote type and quote currency, plus+ the financial currency and underlying symbol when Yahoo declares them. A+ secondary listing no longer reads as the primary one. `GOOGL.BA` is an ARS+ line in Buenos Aires carrying Alphabet's USD financials, and `GGALD.BA` is a+ USD line whose issuer reports in ARS. Nothing is inferred from the ticker.++- **Volume statistics in `technical_indicators`** (#1571, thanks @zeus229). The+ tool returns the latest volume, its 20-bar mean and their ratio, computed from+ the bars it already fetched. The mean needs a complete 20-bar window.+ `volume.unit` is the unit the serving source declares (board lots on the+ A-share sources, shares on the Yahoo family, `null` when undeclared), because+ the two are 100x apart and nothing in the numbers tells them apart.++- **Argentina (BYMA) market data** (#1543, thanks @zeus229). A `.BA` symbol — a+ BYMA listing or a locally traded CEDEAR — is its own market `ar_equity`,+ quoted in ARS, served by `yahoo` → `yfinance` → `local`, and reported as+ market `ar` by `get_stock_profile`. Execution is deliberately not modelled: an+ Argentine backtest raises instead of borrowing US or crypto commissions, lot+ sizes, settlement and short-selling rules. The Web UI carries the market too —Settings source priority, the positions asset-class grouping and the sectorunion — and with it `uk_equity`, `vietnam_equity` and `index`, which had beenmissing from one or more of those three since they landed. `EnvConfig.data`'s@@ -49,68 +72,69 @@ This project adheres to [Semantic Versioning](https://semver.org/).technical-indicator tool canonicalises its documented `1wk` / `1mo` andfetches whole periods instead of two days per bar.-- **IM channels are configured from the Web UI** (#1520, #1529; #1519, #1530).- The Settings panel renders each built-in channel's fields from backend- metadata, masks secrets (`****` plus the last four characters, never sent- back), tests the unsaved form against the provider (`ok` /- `invalid_credentials` / `network` / `unsupported`), and hot-swaps only the- saved channel with no restart. DingTalk and QQ have guided setups; a shared- token probe backs both. Saves write `channels.<name>` in+- **IM channels are configured from the Web UI** (#1520, #1529; #1519, #1530,+ thanks @shadowinlife). The Settings panel renders each built-in channel's+ fields from backend metadata, masks secrets (`****` plus the last four+ characters, never sent back), tests the unsaved form against the provider+ (`ok` / `invalid_credentials` / `network` / `unsupported`), and hot-swaps only+ the saved channel with no restart. DingTalk and QQ have guided setups; a+ shared token probe backs both. Saves write `channels.<name>` in`~/.vibe-trading/agent.json` atomically at mode 0600; a YAML config is shownread-only. Reloads of one channel are serialized in `ChannelManager`, and`stop_all` cancels a reload's start that is still connecting.+- **Email and WebSocket join the guided Web UI channel setup** (#1544, after- #1519). Both channels get hand-written field metadata — localized- labels, masked secrets — plus connection tests and per-channel hot apply.- Email's test probes the real IMAP login, mailbox select and SMTP login- without ever sending a message; WebSocket is a server channel with no- remote credentials, so its test validates the TLS cert/key material and- binds every resolved address locally — an address held by the- already-running server reads as the expected state, not a failure. Saving- a WebSocket config hot-swaps the server: connected clients (including the- Web UI chat) briefly disconnect and reconnect. A hand-written field hint's- secret flag is now authoritative for the keys it covers — the regex- fail-safe still masks every unhinted key — so token-shaped non-secrets- such as `websocket_requires_token` stay visible and editable instead of- being masked, and a form save can no longer silently flip the stored token- requirement. The web config routes now build the WebSocket adapter with- the gateway services it requires; previously every WebSocket save or test- failed validation. Email's polling loop also skips stale-config+ #1519, thanks @shadowinlife). Both channels get hand-written field metadata —+ localized labels, masked secrets — plus connection tests and per-channel hot+ apply. Email's test probes the real IMAP login, mailbox select and SMTP login+ without ever sending a message; WebSocket is a server channel with no remote+ credentials, so its test validates the TLS cert/key material and binds every+ resolved address locally — an address held by the already-running server reads+ as the expected state, not a failure. Saving a WebSocket config hot-swaps the+ server: connected clients (including the Web UI chat) briefly disconnect and+ reconnect. A hand-written field hint's secret flag is now authoritative for+ the keys it covers — the regex fail-safe still masks every unhinted key — so+ token-shaped non-secrets such as `websocket_requires_token` stay visible and+ editable instead of being masked, and a form save can no longer silently flip+ the stored token requirement. The web config routes now build the WebSocket+ adapter with the gateway services it requires; previously every WebSocket save+ or test failed validation. Email's polling loop also skips stale-configdelete/move post-actions once a stop or hot swap begins, while the- already-fetched batch is still delivered. Both the probe and the polling- loop now send an RFC 2971 IMAP `ID` after login: NetEase mailboxes- (163/126/yeah.net) accept the login but reject the first `SELECT` with- `Unsafe Login` until the client identifies itself, so without this the- Email channel could not read the most common Chinese mailboxes. The- identification is static and carries no user data or secrets, and servers- that ignore `ID` are unaffected; verified end-to-end against a live- 163.com account.-- **Bahasa Indonesia** UI locale and `README_id.md` (#1482). The README count- tests cover it, and `MANIFEST.in` now ships `README_es.md` and- `README_id.md`.-- **Portfolio valuation accepts any ISO-4217 currency** (#1510). A source whose- currency has no FX rate fails with the currency named and is excluded from- the totals; the healthy sources stay in.--- **Gildata (恒生聚源) joins the A-share fallback chain as a token-gated- source**. The new `gildata` loader talks to the vendor's raw-api MCP- endpoint (one JSON-RPC POST per call, token in the Authorization header) and- serves A-share daily OHLCV through the `StockDailyQuote` tool with+ already-fetched batch is still delivered. Both the probe and the polling loop+ now send an RFC 2971 IMAP `ID` after login: NetEase mailboxes+ (163/126/yeah.net) accept the login but reject the first `SELECT` with `Unsafe+ Login` until the client identifies itself, so without this the Email channel+ could not read the most common Chinese mailboxes. The identification is static+ and carries no user data or secrets, and servers that ignore `ID` are+ unaffected; verified end-to-end against a live 163.com account.++- **Bahasa Indonesia** UI locale and `README_id.md` (#1482, thanks @Yoruxyv).+ The README count tests cover it, and `MANIFEST.in` now ships `README_es.md`+ and `README_id.md`.++- **Portfolio valuation accepts any ISO-4217 currency** (#1510, thanks+ @zeus229). A source whose currency has no FX rate fails with the currency+ named and is excluded from the totals; the healthy sources stay in.++- **Gildata (恒生聚源) joins the A-share fallback chain as a token-gated source**+ (#1474, thanks @sambazhu). The new `gildata` loader talks to the vendor's+ raw-api MCP endpoint (one JSON-RPC POST per call, token in the Authorization+ header) and serves A-share daily OHLCV through the `StockDailyQuote` tool with`restorationStatus=1` — forward, split-AND-dividend adjusted bars, stamped`split_dividend` in the price-caliber table after being measured against a- live payload. Volume arrives in 万股 and is converted to shares- (`volume_units: shares`); `avgprice`/`prevcloseprice` are ignored because- the vendor keeps them on a different adjustment basis than the adjusted- OHLC. An unresolvable symbol answers `rows: []` (never an error), so the- chain keeps walking. Auth is `GILDATA_TOKEN` (Settings page field or env;- `GILDATA_BASE_URL` overrides the endpoint, `VIBE_TRADING_GILDATA_MIN_INTERVAL`- the 0.3s default spacing). Without a token the loader reports unavailable- and the chain skips it — no behavior change for existing users beyond one- reorderable entry at the chain's tail. One migration note: a- `MARKET_DATA_ORDER_A_SHARE` value saved before this change is a permutation- of the old 7-source chain, so it stops validating once `gildata` joins —- the Settings card flags it and the default order applies until the saved- order is re-saved (one click) with the new source included.+ live payload. Volume arrives in 万股 and is converted to shares (`volume_units:+ shares`); `avgprice`/`prevcloseprice` are ignored because the vendor keeps+ them on a different adjustment basis than the adjusted OHLC. An unresolvable+ symbol answers `rows: []` (never an error), so the chain keeps walking. Auth+ is `GILDATA_TOKEN` (Settings page field or env; `GILDATA_BASE_URL` overrides+ the endpoint, `VIBE_TRADING_GILDATA_MIN_INTERVAL` the 0.3s default spacing).+ Without a token the loader reports unavailable and the chain skips it — no+ behavior change for existing users beyond one reorderable entry at the chain's+ tail. One migration note: a `MARKET_DATA_ORDER_A_SHARE` value saved before+ this change is a permutation of the old 7-source chain, so it stops validating+ once `gildata` joins — the Settings card flags it and the default order+ applies until the saved order is re-saved (one click) with the new source+ included.- **A Robinhood account can be a read-only portfolio source** (#1428). The new`robinhood-live-mcp-readonly` profile uses the same MCP server and OAuth@@ -125,6 +149,97 @@ This project adheres to [Semantic Versioning](https://semver.org/).that also holds options, crypto, futures, event contracts, mutual funds orfixed income fails the source instead of showing an equity-only view.+- **A weekly health check of every keyless data source** (#1627, closes #1623,+ thanks @he-yufeng) — `python -m backtest.loader_health` fetches one liquid+ symbol from each of the 15 keyless network loaders, each in its own process+ with a temporary home, no inherited keys and the loader cache off, and+ requires normalised OHLCV whose last bar is at most 14 days old. The+ `Public loader health` workflow runs it every Monday and on demand. An+ unreachable endpoint, a missing dependency, malformed bars and a stale+ series all fail the lane; as first proposed, an unreachable endpoint counted+ as a skip. An offline test fails when a new keyless loader has no canary+ symbol.++- **OpenCode is a built-in provider, and each request carries the session id+ OpenCode Go requires** (#1416, closes #1415, thanks @tonydo) —+ `LANGCHAIN_PROVIDER=opencode` with `OPENCODE_API_KEY` / `OPENCODE_BASE_URL`,+ in Web Settings and both CLI onboarding flows. OpenCode Go refused every call+ without `x-opencode-session` (400 `MissingSessionID`); the adapter now sends+ the run's session id to any `opencode*` provider or `opencode.ai` host, and+ the compaction summary call, made on its own thread, carries the same id+ (`fd6c7c7c`) instead of reaching OpenCode as a second conversation.++- **KIS (Korea Investment & Securities) is a connector for KOSPI and KOSDAQ+ equities** (#1407, thanks @as950118) — paper (모의투자) and live sit on+ separate hosts with separate OAuth tokens, so the paper-trade profile sends+ orders to KIS's own paper server rather than a local simulator. Live is+ read-only: no `kis-live-trade` profile exists, and `place_order` /+ `cancel_order` refuse a non-paper config before anything else. A finishing+ commit (`970c235f`) pins the paper sell and cancel TR_IDs, which no test+ covered, and makes a balance or order walk that hits its 20-page cap raise+ instead of returning a short position list that reads as complete.++- **Upbit is a KRW crypto connector, capped at paper plus read-only** (#1408,+ thanks @as950118) — one key pair reads the same account whatever profile is+ declared, so there is no runtime paper/live discriminator and no live order+ path. Paper orders are simulated against a live Upbit quote: no quote means+ no fill, a limit order fills only if immediately marketable, and cancel+ acknowledges only ids the simulator issued. Needs+ `pip install "vibe-trading-ai[upbit]"`.++- **Toss Securities is a read-only connector for Korean and US equities**+ (#1409, thanks @as950118) — Toss documents no sandbox and no paper/live+ discriminator, so, like Trading 212, it refuses every order and cancel,+ paper included. Reads follow the published schema (`/api/v1/holdings`,+ `/api/v1/orders`); closed-order history is paged to completion, and a walk+ past its cap, or a `hasNext` page with no cursor (`b5ff36ba`), is an error+ rather than a silently truncated list.++- **Scalable Capital is a read-only MCP connector that declares quotes only**+ (#1417, thanks @cgycorey) — Scalable has no paper account and publishes tool+ names without argument schemas. `scalable-live-mcp-readonly` withholds+ `account.read` / `positions.read` because nothing maps the holdings reply+ yet: declaring them made it a portfolio source whose refresh stored a+ complete snapshot with zero positions. The 29 published tools are pinned+ literally as 18 read and 11 write (unknown names resolve to write), and a+ hand-enabled order tool is wrapped by the live order guard and refused.++- **The README's broker capability matrix is generated from the profile+ registry** (#1629, thanks @he-yufeng) — one row per built-in profile (55+ across 18 connectors): environment, transport, read-only or write-enabled,+ declared reads, and whether placement requires a mandate. A test fails CI+ when the block and the registry disagree; regenerate with+ `PYTHONPATH=agent python -m src.trading.capability_matrix`. As merged it+ folded each connector into one row; `dec2c0e6` renders each profile on its+ own, so a paper order capability never reads as a live permission, and+ says the rows are declarations, not runtime verification.++- **Run cards record the backtest's execution and cite metrics against the+ saved CSV** (#1612, closes #1610, thanks @lorenzozanee) — `run_card.json`+ (schema `1.0`) carries a trace with UTC start and end, status, and SHA-256+ hashes of the config and the metrics, never the payloads. As merged, every+ scalar metric was cited whenever `artifacts/metrics.csv` existed; after+ `0244ecea` a metric is cited only when its value matches the file's single+ data row and the file matches its manifest checksum. The Markdown card and+ the Run Detail page show both sections.++- **A read-only extraETF export reader** (#1406, thanks @cgycorey) — the+ Investments export becomes positions and the Transaktionen export typed+ movements, so a `Dividende` row is never read as a holding and a+ transactions file cannot pass as a portfolio. It fails closed on anything it+ cannot read without guessing (unknown header or label, US-ordered numbers,+ more than 8 decimals, one instrument twice under one Portfolio ID). Nothing+ calls it yet; the portfolio container side is still open under #1170.++- **Strategy-store adaptation substrate** (#1391, thanks @modelpath-dev) — a+ child strategy points at its parent through `derived_from` (SQLite schema 3,+ an additive nullable column), and `AdaptationPatch` can change only+ universe, name and position sizing, never `signal_definition`. The child+ starts with no evidence and `UNVALIDATED`; `a7bd2e30` also clears the+ parent's validator, approver and versions, and makes `register_adaptation`+ require the parent's evidence, refusing a stale or insufficient parent.+ Nothing outside tests calls it yet (#1149 stays open).+### Changed- **`source="qveris"` refuses markets with splits and dividends** (#1494). It@@ -135,22 +250,27 @@ This project adheres to [Semantic Versioning](https://semver.org/).Both QVeris budget gates reserve only a flat per-call quote now:`1 credits/result` had reserved 1 credit for a call that billed 9.66, and`qveris_execute` returns `quote_not_bounded` for any other shape.-- **A price index is stamped `raw`** (#1541). Eastmoney's fqt=1 equals fqt=0- on every bar of the A-share indices measured, Tencent serves only `day` for- the SSE/SZSE ones, and Yahoo's adjclose equals close on ^GSPC / ^NDX / ^HSI, but the- per-source table stamped an A-share index additive and a Yahoo index++- **A price index is stamped `raw`** (#1541). Eastmoney's fqt=1 equals fqt=0 on+ every bar of the A-share indices measured, Tencent serves only `day` for the+ SSE/SZSE ones, and Yahoo's adjclose equals close on ^GSPC / ^NDX / ^HSI, but+ the per-source table stamped an A-share index additive and a Yahoo indexdividend-adjusted. `price_caliber` takes the symbol; A-share index codes arematched per exchange, since 000001.SZ is Ping An Bank.-- **Portfolio valuation version 3** (#1510): snapshots stored under version 2- are no longer read, so portfolio history starts again after upgrading.++- **Portfolio valuation version 3** (#1510, thanks @zeus229): snapshots stored+ under version 2 are no longer read, so portfolio history starts again after+ upgrading.+- **Tencent, Eastmoney and AKShare A-share prices are labelled- `split_dividend_additive`** (#1497, #1493). Their forward adjustment- subtracts cash dividends from the level instead of scaling by a ratio (on- 600519.SH, qfq minus raw takes five values over 500 bars against 367+ `split_dividend_additive`** (#1497, #1493, thanks @cgycorey). Their forward+ adjustment subtracts cash dividends from the level instead of scaling by a+ ratio (on 600519.SH, qfq minus raw takes five values over 500 bars against 367distinct ratios), so a basket mixing them with multiplicative sources warns,and a run served only by them warns that its returns are not total returns.Tencent's HK series is labelled `raw`: it serves no adjusted HK bars. Theprices themselves are unchanged.+- `connector account` renders Binance spot balances as Asset / Free / Locked /Total, Futu's per-currency `assets`, and Trading 212's `cash` and `metadata`(#1539); before, Binance printed blank rows and the other two printed "No@@ -161,27 +281,28 @@ This project adheres to [Semantic Versioning](https://semver.org/).percent sign, a currency mark or a table cell makes it a measurement; dates(including year-less "09-14"), years, security codes, line-leading ordinalsand code in a language-tagged fence are structure; a plain integer is checked- only as a price of an instrument quoted in the thousands — so a sentence- gets the same verdict in every language. A measurement that equals a price- or volume this session's tools returned needs nothing more. Any other one is+ only as a price of an instrument quoted in the thousands — so a sentence gets+ the same verdict in every language. A measurement that equals a price or+ volume this session's tools returned needs nothing more. Any other one isdeclared in a fenced `figures` block at the end of the answer, one line perfigure — `value | role | note | ref` — and checked by role: `observed` must- appear in the referenced call or tool (`ref` may name either), filtered to- the figure's own symbol, with a currency-marked figure compared only against- money-denominated values and a percent only against non-price ones;- `derived` needs an evaluable formula whose added or subtracted operands are- all observed, matching the figure at the precision and sign it was written- with; `proposed` is a price inside its own symbol's observed range or a- derivation; `cited` must name its source on the figure's own line, because- the block is stripped before anyone reads the answer; `count` covers counts,- weights, thresholds and probabilities, while a currency-marked figure, or one- inside its instrument's price range that no derivation uses as a factor, is- checked as observed when declared `count`. No role other than `observed` may sit in a+ appear in the referenced call or tool (`ref` may name either), filtered to the+ figure's own symbol, with a currency-marked figure compared only against+ money-denominated values and a percent only against non-price ones; `derived`+ needs an evaluable formula whose added or subtracted operands are all+ observed, matching the figure at the precision and sign it was written with;+ `proposed` is a price inside its own symbol's observed range or a derivation;+ `cited` must name its source on the figure's own line, because the block is+ stripped before anyone reads the answer; `count` covers counts, weights,+ thresholds and probabilities, while a currency-marked figure, or one inside+ its instrument's price range that no derivation uses as a factor, is checked+ as observed when declared `count`. No role other than `observed` may sit in aprice column. The block is stripped from the released answer and never- streamed; the artifact keeps it. The price-word, level-word,- derivation-phrase and metric-subject catalogues are deleted with it- (65 compiled regexes → 12, all shape), and `src/agent/grounding.py` is now- the `src/agent/grounding/` package.+ streamed; the artifact keeps it. The price-word, level-word, derivation-phrase+ and metric-subject catalogues are deleted with it (65 compiled regexes → 12,+ all shape), and `src/agent/grounding.py` is now the `src/agent/grounding/`+ package.+- **A rejected answer costs one correction round, not four, and is thenreleased with its failing figures cut rather than refused.** The correctionprompt lists every failing figure as written, with its declared role and what@@ -198,16 +319,59 @@ This project adheres to [Semantic Versioning](https://semver.org/).budget. The run stays `degraded` with a reason naming the redaction, therelease path's rechecks are not counted as rejected drafts, and the shippeddocument is recorded under `released` in `artifacts/grounding_evidence.json`.+- **A draft missing only a source or currency word gets a data note appended,not another model round.** The canonical symbol is deliberately not repairedthis way: it is the figure's subject, and a note naming it under an answerabout another instrument would footnote a misattribution.+- **The chat shows that the answer is being checked.** The agent loop emits`grounding_status` (`{stage: "revising", round, issues}` after a rejection,`{stage: "released_redacted", removed}` before a cut release), and the chatpage shows "Checking the figures in this answer (round N)…" until answer textarrives, in all eight locales.+- MetaTrader 5 no longer picks a broker alias when several match+ (`47a036d6`, with #1597 and #1598). Quotes, history, orders, symbol search+ and the backtest loader took the shortest suffixed name (`EURUSDm` over+ `EURUSDz`); they now refuse and ask for the exact native symbol. The+ connector still tries a configured `symbol_suffix` first; the backtest+ loader does not read it.++- `run_card.json` `schema_version` is `1.0` (was `0.1`), with the optional+ `tool_traces` and `citations` keys from #1612.++- Broker connectors: 18, up from 14 (KIS, Upbit, Toss Securities, Scalable+ Capital). `test_readme_counts.EXPECTED_BROKERS` is checked against the+ profile registry, not incremented by hand.++- The opt-in loader cache is at version 7 (#1411, thanks @shadowinlife). Entries+ written under version 6, which include Tencent series the forward walk+ truncated, are never matched again and are fetched afresh.++- `VIBE_TRADING_BAOSTOCK_TIMEOUT` (default 30 seconds) bounds each BaoStock lock+ wait, connection and complete message (#1615, thanks @he-yufeng). It is not a+ deadline for a whole multi-symbol fetch.++- The `a_share` fallback chain is `tencent`, `mootdx`, `eastmoney`,+ `baostock`, `akshare`, `tushare`, `gildata`, `local` (#1474, thanks @sambazhu).++- Factor values change for 23 alphas, so an IC, bench or comparison stored+ before this release is not comparable with one computed after it:+ `qlib158_cntp*` / `cntn*` / `cntd*` (15 factors; the first value lands one+ bar later and windows touching a gap are NaN, #1459), `qlib158_wvma5` …+ `wvma60` (5; every value, #1594), and `gtja191_103` / `133` / `177` (3;+ direction and scale, #1604). Patell z for `market_adjusted` and+ `mean_adjusted` event studies moves slightly as well (#1449, thanks @Shizoqua).++- **A grounding correction turn is offered no tools** (#1600, thanks @zeus229)+ — after a rejected draft the model was handed every tool again, so revising+ a figure could restart research. Unless the gate asks for recovery (an+ unresolved symbol or a missing price), the correction turn gets none. A call+ made anyway is not run and, since `47a036d6`, spends one of the two+ correction rounds; at the cap the last real draft goes through the existing+ cut-and-recheck release.+### Removed- **The Requesty built-in provider.** Its capabilities were OpenRouter's, with@@ -217,231 +381,159 @@ This project adheres to [Semantic Versioning](https://semver.org/).### Fixed-- **A dollar-quoted BYMA or TSX line, or a Hong Kong RMB or USD counter, no- longer enters a single-currency backtest** (#1576, after #1543). Both venues- list USD lines beside their home-currency ones. On 2026-09-24 Yahoo declared- USD for GGALD.BA (4.20), SPYD.BA and AAPLD.BA, and for DLR-U.TO and- XUS-U.TO. `ar_equity` and `ca_equity` are each one static-currency pool, so- such a line was priced as pesos or Canadian dollars. As on the LSE, the- loaders now read the declared currency. They admit a `.BA` line only in ARS- and a `.TO` / `.V` line only in CAD. A trailing `D` does not decide it:- YPFD.BA is a peso line. Every other declared quote currency is recorded in- market-data provenance, and the grounding gate asks an answer to name that- currency rather than the one the suffix implies. Hong Kong is decided by the- code, because the Hong Kong chain starts with sources that declare no- currency: HKEX's Stock Code Allocation Plan assigns 80000-89999 to products- traded in renminbi and several 09xxx ranges (plus 10900-10999 and- 41500-41599) to products traded in USD. So an RMB counter (80700.HK, 375.40- CNY on 2026-09-24) or a USD ETF (9834.HK) is refused in a Hong Kong or- cross-market backtest, is counted in its own currency by the mixed-currency- guard, and is quoted in that currency by the grounding gate, whichever- source served it. The table matched the currency Yahoo declared for all 24- codes probed across the ranges.-- **Grounding identity and symbol search know every market the data layer- routes** (#1565, #1575). `.BA`, `.L` and `.VN` symbols had no- canonical-symbol scan, venue or currency in grounding. Symbol search- labelled Indian, Korean, British, Vietnamese and Argentine results `global`.- A parity test now takes the market list from the backtest's own currency- table. An answer may write £, ₫ or AR$ for the currency it has to name.-- **An explicit provider header survives an ambient twin** (#1573, #1568).- When `OPENAI_CUSTOM_HEADERS` names a header the provider also sets (a- `user-agent` beside the Kimi / NVIDIA / OpenCode `User-Agent`), openai 3.19.2's- case-insensitive merge dropped the provider's value. That turned CI red for- every PR on 2026-09-24. The first fix proposed would have sent both headers- under openai 2.53, the version the lock file and the Docker image install.- Every ambient spelling is now omitted and the explicit value re-set after- the omits, which holds under both.-- **Backtests align calendars at any timestamp resolution** (#1560). The engine- merged symbol calendars and placed prices through raw nanosecond integers,- but a local duckdb source serves microseconds. Such a run was dated- 1970-01-21, and beside a nanosecond source one symbol's closes came back- empty. The rebalance mask had the same unit mismatch.-- **A research goal that is an order is still refused** (#1562). The execution- filter no longer rejects research that mentions shares or coins ("should the- fund sell its GOOGL shares"). It refuses an objective that opens with buy or- sell, puts a quantity right after the verb, or writes 买入 / 卖出 with a- quantity and a unit. The first version of the change had made "Buy 100- shares of NVDA" an accepted goal.-- **A strict-bench OOS split needs two IC observations on each side** (#1559).- A split outside the loaded prices is refused. An alpha whose IC series- leaves fewer than two observations on either side is skipped, with the- counts in the reason. The IC series ends one forward-return horizon before- the prices, so a split one bar from the end used to publish verdicts- measured on nothing.-- **Quant inputs are checked before they are used** (#1555-#1558). HRP aligns a- supplied correlation matrix, and the covariance's own rows, to the- covariance's column labels; swapping two rows used to move the weights with- no error. Purged cross-validation refuses an unordered or duplicated- timestamp index, and a label that ends before it starts. Impact models and- every fixed-income entry point refuse NaN and infinities instead of- returning NaN prices, durations and curves.-- **Shadow-account overtrading uses the whole trading window** (#1563). A long- hold that closed before later short trades no longer shrinks the span.-- **Asset growth is year over year on the daily panel** (#1564). Derived- fundamentals run on a panel densified to the price calendar, where a one-row- lag compared yesterday with today: 0% on ordinary days and a one-day spike- at each filing. The lag is now 365 days, as-of.+#### Long research runs stop losing what they fetched++A user's two runs on the Codex provider (GPT-6 Sol) failed as "0 steps ·+no_progress": a risk-parity backtest after 1m03s and a comparison of A-share+and Hong Kong game companies after 4m50s. They were two different failures.++- **Context compaction follows the model's real context window** — compaction+ fired at a fixed estimate of 40,000 tokens (characters / 4 of the messages,+ not counting tool definitions), while the system prompt alone is about+ 12,500 of them. From the third step it cleared every tool result but the+ last three, so a comparison across a dozen companies re-fetched the same+ filings until the no-progress stop: reproduced on v0.1.15 with DeepSeek V4+ Pro, seven A-share issuers' filings fetched four to five times each, stop at+ iteration 16; with a larger budget the re-fetches went from 54 to 0. The+ window now comes from `VIBE_TRADING_CONTEXT_WINDOW`, else a limit learned+ from the provider's own context-length error during the run, else+ `src/providers/context_windows.json` (272K for the Codex GPT-6 / GPT-5.6+ models, 1M for DeepSeek V4, each entry naming its source), else 128K, and+ `VIBE_TRADING_CONTEXT_MAX_TOKENS` (200K) caps what a prompt may cost. Usage+ is read from the input tokens the provider reports. The three layers fire at+ 80 / 90 / 100% of what the system prompt and tool schemas leave, and clear+ the oldest results first, only down to the line. A context-length error is+ compacted and retried instead of failing the run. `TOKEN_THRESHOLD` is+ deprecated and ignored with a warning; remove it from `.env`.++- **`read_file` says where the agent may read, and lists a directory** — the+ 1m03s failure: on the same prompt, a live DeepSeek V4 run went looking for+ the backtest engine's source (`../../agent/src/backtest`, the checkout's+ absolute path, `skills/../backtest/engine.py`) and every attempt got the same+ "File not found or path escapes workspace". Each spelling was a new call and+ none an observation, so eight rounds exhausted the no-progress budget in 93+ seconds, before any backtest ran. The refusal now names the case+ (`outside_readable_roots` or `not_found`), the readable roots (the run+ directory and `skills/`) and where the engine's contracts are documented+ (`load_skill`); a directory inside those roots returns its entries. The+ boundary itself is unchanged.++- **A failed or cancelled run keeps its steps** — only a completed run saved+ its tool trail, so the Web history showed "0 steps" for exactly the runs+ someone would want to inspect.++- **A no-progress stop says which calls produced nothing, and why** — each+ failing or repeated call is listed with its reason and count ("read_file+ failed (… not found) ×3"), instead of a message about previous summaries+ and artifact paths written for a different failure.++- **`get_market_data` fails when no requested symbol returned data** — an+ envelope holding only `_unresolved` was a success, so a run whose sources+ were all down counted it as progress. It is now `error_code:+ no_market_data`; a partial result is still a success, and grounding still+ records every symbol as unavailable.++- **Codex (ChatGPT sign-in) keeps the system prompt, its reasoning and its+ call ids** — a system message the loop adds mid-run (a retry hint, a+ write-the-file reminder) replaced the whole 49,000-character system prompt+ for every later request; it is now sent as a developer message, as the+ Codex CLI does. Reasoning items are carried from turn to turn the way the+ Codex CLI replays them, and resent without them if the backend refuses;+ tool calls keep the item ids the backend issued instead of a fabricated+ `fc_0`; the prompt-cache key is stable for the session instead of changing+ on every request. The default Codex model is `gpt-6-sol`: `gpt-5.4` is+ refused for ChatGPT accounts.++#### A backtest report is checked against the backtest++A risk-parity versus equal-weight report ended in the canned refusal in 12 of+15 replayed runs. Replaying all 26 stored drafts through the gate (880 of 882+recorded rejections reproduced) showed that four in five refused figures were+values the backtest itself had written.++- **A backtest's own output grounds the report about it** — only metrics+ whose names the gate knew (Sharpe, return, drawdown, volatility, win rate)+ were evidence. Every scalar of the run card, the metrics row, the risk+ X-ray, the Monte Carlo validation and the rebalance summary now is (lists+ such as the 1,000 simulated Sharpes are not descended, and a file other than+ the metrics must match the run card's SHA-256 manifest). A per-bar table+ counts only for the rows a `read_file` showed, while it is still byte for+ byte what the engine wrote; a file the model wrote is never evidence. A+ figure may name the run it came from — `rp`, `rp/artifacts/metrics.csv`,+ `risk_parity target_positions.csv`, `rp::sortino` — and a value from one+ backtest does not ground a figure declared from another. Replayed:+ `value_mismatch` 196 → 1, `not_in_referenced_call` 167 → 7. On the same+ runs a real value declared with its run passes 97% of the time and the+ same value moved 2–10% passes 0.7%; the undeclared path was deliberately+ not widened (the whole metrics row there would have passed 16% of invented+ values).++- **A comparison's arithmetic is read as it is written** — a formula glued to+ its labels ("等权 Sortino 1.2115 − 风险平价 1.1329") or followed by a+ bracketed aside was "not arithmetic"; an operand written as a percent was+ read as a whole number, so "13.28% − 12.98%" anchored on nothing and+ "0.666 × (1 − 3%)" computed the wrong value; a sum of squared weights+ (HHI) could not be written at all; a formula's negative operand (a+ drawdown) could never anchor; and a field two runs both hold anchored+ neither. All now evaluate (only a literal square or cube is accepted as a+ power), and the ones that still fail are the model's own arithmetic.+ A declared value may be a fraction (`1/3`) or a multiple (`5.5x`).++- **A report that still fails after a completed analysis is released with the+ failing figures cut** — naming a symbol made any request a market answer,+ and a market answer with no fetched price was never released cut. A+ completed analysis now is; a price question with neither prices nor an+ analysis still gets the refusal, and the refusal after an analysis no+ longer says the draft cited prices.++- **A draft ending in DeepSeek's closing tool-call tags is caught** —+ `</||DSML||invoke>` slipped past the pattern for opening tags and reached+ the gate as unreadable declaration lines.++- **A correction says what went wrong with a figure the model did declare** —+ `0.3897` and `38.97%` stay two assertions, and so do `−1.51pp` and a formula+ that computes `+1.51`, but the model is now told which value it declared+ (also when it has the opposite sign) and that its formula runs the other way+ round, instead of "not declared" or "its own note evaluates to 1.51%". In a+ live run after the evidence fix, eight differences were still cut because+ the second draft rewrote the numbers rather than the formulas. The+ correction also asks for the corrected answer alone, in the user's+ language: that draft had opened with "The rejection was because…", in+ English to a user writing Chinese, and the sentence was released.++#### The grounding gate+- **A tail-risk figure names its own field once a session holds more than one**- (#1425, after #1444). A call- or tool-scoped ref (`ref x1`) pooled every- tail-risk field that call returned, and an undeclared tail-risk percent pooled- every tail-risk value in the session, so an ES 95% could quote the VaR 95%- value and pass. The rule is read off the evidence, never off prose:- `tail_risk_identity()` takes (measure, confidence) from the field name a tool- returned, so `var_95` and `es_95` are two identities while `cvar_99` and- `es_99` are one. When a scope holds two or more and the figure's value matches- one of them, it is refused with `tail_risk_needs_field_ref`, whose correction- names every identity the session holds; a figure matching no tail-risk value- keeps its old reason, so a fabricated number is still a mismatch rather than a- missing ref. Accepted cost, stated in the issue: an undeclared VaR 95% beside a- VaR 99% is correct today and now costs one correction round. The system prompt- states the rule so a risk report declares the ref on its first attempt.-- **Mean-variance style optimizers size a short by its own expected return**- (#1548). The optimizers scored every position as if it were held long, so the- strongest shorts (most negative drift) looked worst and got the least capital.- `BaseOptimizer.optimize` now negates a short's column before building the- context, so `mu` is the position's expected return AND `cov` is the position- covariance `D Σ D`: a long and a short of two +0.92 correlated names hedge- instead of reading as correlated (the old objective put the whole book on the- long, the new one splits it 0.514 / -0.486). Risk-parity and- max-diversification contexts get the position covariance with it;- volatility-only contexts are unchanged, since std(-r) == std(r).-- **Monte-Carlo validation annualises at the venue's bars per year** (#1546). It- hardcoded 252 while `bootstrap_sharpe_ci` and `walk_forward_analysis` in the- same report used the resolved value, so a crypto (365) or forex (260) run- published two disagreeing Sharpes.-- **A futures order on a negative-price bar is no longer rejected outright**- (#1547). `FuturesBaseEngine` dropped the `abs(price)` its base class relies on- under `allow_nonpositive_prices`, so size came out negative and the caller's- `size <= 0` guard refused every order.-- **A factor's IC ratio is unset for a negative baseline** (#1549). Dividing two- negative IC means gives a positive ratio, so a rolling IC that got further- below zero reported as improvement. The signal itself took the worst of the- available metrics and was not fooled; the published number was.-- **Two memories sharing a title keep separate index rows and links** (#1545).- Under `VT_MEMORY_HIERARCHY` an entry lives at `{memory_type}/{slug}.md`, so- two entries with one title had the same `path.name`: the second `add()`- overwrote the first's `MEMORY.md` row, and the semantic-link block beside it- excluded the other entry as "self" and wrote an ambiguous target. Both key on- the path relative to the memory dir now; a sidecar written earlier, with a- bare filename or an absolute path, is still read.-- **The Email channel verifies TLS certificates by default** (#1544). The- implicit-SSL paths (`IMAP4_SSL` / `SMTP_SSL`) in both the connection probe- and the polling/send adapter used Python's default unverified context- (`CERT_NONE`, no hostname check), so the mailbox password travelled to a- server whose certificate was never checked — an active network attacker- between the operator and their mail provider could harvest it. All four call- sites now build their context through a shared `email_tls_context()` that- verifies the certificate and hostname against the system CA bundle, matching- the STARTTLS path, which already did. A new `verify_tls` field (default- `true`, surfaced in the guided Email setup in all nine locales) is the- documented opt-out for self-signed or internal-CA servers. **Upgrade note:**- a mail server with such a certificate connected before this change and now- fails the TLS handshake until `verify_tls` is set to `false` for it.-- **Plain IMAP no longer sends the mailbox password in clear text** (after- #1544). With `imap_use_ssl` off, the Email channel and its connection test- logged in over an unencrypted connection, and the guided setup put that- switch one click away. A new `imap_use_tls` (on by default, the IMAP twin of- `smtp_use_tls`) upgrades the connection with STARTTLS before `LOGIN`; a- server that offers no STARTTLS is reported as a network failure before the- password is sent. `verify_tls` now governs STARTTLS on IMAP and SMTP as well- as implicit SSL, so a self-signed server can opt out on any path. **Upgrade- note:** an `imap_use_ssl: false` setup against a server without STARTTLS- stops connecting until `imap_use_tls` is set to `false` for it, which sends- the password in plain text again.-- **Settings-write routes reject cross-site browser requests** (#1544).- `require_settings_write_auth` did not apply the cross-site guard that its- siblings `require_auth` and `require_event_stream_auth` enforce on unsafe- methods, so a malicious web page could attempt a CSRF write against any- settings route — including the new `POST /channels/email/test`, which merges- stored credentials with a caller-supplied patch and could be steered to send- the stored mailbox password to an attacker-chosen host. The guard now runs- first on every settings-write route; same-origin Web UI calls and- non-browser clients (CLI/curl, no `Origin`) are unaffected.-- **Read-only results lost to context compaction are restored, not refetched**- (#1488). A successful read-only call whose payload compaction removed is- replayed from the run's own cache, at most six times a run; past the cap it- runs again, as it did before, instead of being refused with "use the- previous result" for a payload the model cannot see. The cap is checked per- restore, and any successful write empties the cache, so a regenerated file- is read afresh. `read_url` opts in; `no_cache=true` always fetches.-- **Tail-risk figures keep their identity under a field ref** (#1444). A- declared `ref` may name an evidence field by its full path+ (#1425, after #1444, thanks @he-yufeng). A call- or tool-scoped ref (`ref x1`)+ pooled every tail-risk field that call returned, and an undeclared tail-risk+ percent pooled every tail-risk value in the session, so an ES 95% could quote+ the VaR 95% value and pass. The rule is read off the evidence, never off+ prose: `tail_risk_identity()` takes (measure, confidence) from the field name+ a tool returned, so `var_95` and `es_95` are two identities while `cvar_99`+ and `es_99` are one. When a scope holds two or more and the figure's value+ matches one of them, it is refused with `tail_risk_needs_field_ref`, whose+ correction names every identity the session holds; a figure matching no+ tail-risk value keeps its old reason, so a fabricated number is still a+ mismatch rather than a missing ref. Accepted cost, stated in the issue: an+ undeclared VaR 95% beside a VaR 99% is correct today and now costs one+ correction round. The system prompt states the rule so a risk report declares+ the ref on its first attempt.++- **Tail-risk figures keep their identity under a field ref** (#1444, thanks+ @he-yufeng). A declared `ref` may name an evidence field by its full path(`data.tail_risk.var_95`) or trailing part (`var_95`), or one call's field- (`q1::historical_var`); a VaR 99% quoting the 95% value, or ES quoting VaR,- is rejected under such a ref. A field two calls returned with different- values is ambiguous and the correction names the refs to choose from. A- claim scoped to a whole call or not declared is still matched against every- tail-risk value it can see (#1425 stays open).-- **A backtest aborted when a funding debit left cash below zero** (#1542).- CompositeEngine and CryptoEngine subtract crypto funding from capital with- no floor, and the next open then fitted at no scale, not even an empty plan:- the open-basket search kept the full-scale plan and the run died with- "planned order … exceeds available capital", and a rebalance raised- "insufficient capital for position rebalance". Such an open is now skipped- and reported once as `insufficient_capital`, the bar's reductions still- run, and a close whose loss exceeds its margin still aborts the bar (#1274).-- The grounding gate reads decimal commas (#1517): `17,93 %` and+ (`q1::historical_var`); a VaR 99% quoting the 95% value, or ES quoting VaR, is+ rejected under such a ref. A field two calls returned with different values is+ ambiguous and the correction names the refs to choose from. A claim scoped to+ a whole call or not declared is still matched against every tail-risk value it+ can see (#1425 stays open).++- The grounding gate reads decimal commas (#1517, thanks @zeus229): `17,93 %` and`1.410,00 CNY` ground against tool results, two unspaced numbers such as`1400,1777` stay two numbers, and a written figure must match its evidencewithin half a unit of its last written digit (a truncated `30.20%` for30.2052% is now sent back for correction). One shape regex added (14 → 15).-- Block trades, margin trading and financial statements no longer read- Eastmoney rejecting a stale query (code 9501) as an empty result; all six- datacenter callers share `eastmoney_client.datacenter_rejection`.-- **24 alphas emitted a value computed from a missing bar** (#1463, #1523,- #1534, #1452). A comparison over an operand that is NaN because its window- holds the gap is False, and the #1463 sweep, which nudged the bar by a tick,- could not see that. A second oracle compares the gapped run with the gap-free- one. Comparison gates and `np.fmin`/`np.fmax` now mask on their inputs' reach- (`factors.base.observed_over`) rather than on their operands' NaN, so a- correlation undefined on complete data keeps its old verdict; the six- fmax/fmin alphas converted on 2026-09-18 had lost up to 87% of their- post-warmup values in a 5-symbol universe that way and are restored. On- gap-free data no finite value changed.-- Lockup expiry: `free_shares` is the unlocking quantity- (`CURRENT_FREE_SHARES`), not the float; units are stated (万股 / 万元) (#1513,- #1501). The report is `RPT_LIFT_STAGE`; the old one rejected six columns.-- Dragon-tiger seats come from the buy/sell detail reports, surface a provider- rejection, and rank within each listing reason (#1512, #1502).-- Shareholder-count history returns real periods from `RPT_HOLDERNUM_DET`,- states each row's previous period, and warns when the average-holding join- fails (#1518, #1503).-- QVeris resolves adjusted fields as a complete set, one family per response,- and no longer gives an unadjusted bar the adjusted volume (#1527, #1494).-- Channel adapters log through stdlib again: `{}` placeholders, napcat's 17- sites and loguru-only `.opt()` calls lost their messages (signal's- `_safe_handle` raised instead of swallowing) (#1533, #1531). Slack DMs from- unapproved senders get a pairing code (#1524); Discord no longer pings- `@everyone` or roles (#1537); duplicate-reply suppression covers integer- message ids and is bounded (#1476); the WebSocket channel starts again (from- #1521).-- Shadow account: an `other`-market rule no longer matches every symbol- (#1538), and overtrading PnL no longer double-counts trades already- explained (#1475).-- Quantlib: comps refuses negative bridge magnitudes like the DCF bridge- (#1535); accrued interest is exact on coupon dates under 30/360 and rejects- an unknown day count there too (#1509); Kyle's lambda fits an intercept- (#1477).-- ML strategy: a single-class training window is skipped and the previous- model keeps predicting (#1522). Memory: two entries with one title and- different types keep separate index rows, and `forget` names the type when- ambiguous (#1525). Swarm run and task files use unique temp names (#1536).-- `/agent.json`, `/agent.yaml` and `/agent.yml` are gitignored at the runtime- root; the Feishu QR login already wrote channel credentials there.-- **OpenAI's gpt-5.6 models run with their reasoning on the agent's default- configuration** (#1473). `/v1/chat/completions` refuses function tools for- `gpt-5.6-terra` and `gpt-5.6-luna` unless `reasoning_effort` is `none`,- including when no effort is configured and the model applies its own- default, so a plain `LANGCHAIN_PROVIDER=openai` setup failed on its first- tool call with `Function tools with reasoning_effort are not supported`.- The adapter now treats that refusal the way it treats a rejected- `temperature` or `stream_options`: the request is retried on- `/v1/responses`, where the effort travels as `reasoning.effort`, and the- model is remembered for the rest of the process so later calls go there- first. Nothing changes for a model whose chat endpoint accepts tools, and- `LANGCHAIN_USE_RESPONSES_API=true` still selects the route up front and- skips the one failed request.- **A redacted answer keeps its row numbers and its labels** (#1471). When thegrounding gate released a draft with its unverified figures cut, a rankedtable came back with `(omitted※)` in place of 1, 2, 3, of the `12m` in its@@ -455,80 +547,28 @@ This project adheres to [Semantic Versioning](https://semver.org/).or currency figure, and every cell under an OHLC column stay measurements;and the sweep never keys on a single digit. An integer price of aninstrument quoted in the thousands is still checked inside a worded cell.-- **One cash-starved open is one rejection, not twenty-six** (#1470). In- `position_adjustment="hold"`, the search that scales a basket down to the- cash available re-planned every open on each step, and every step whose- size rounded to zero booked a `zero_size` rejection for the same symbol on- the same bar — one contract the cash could not hold read as 26 lot-rounding- failures. The trial plans are silent now; a sleeve that was a real order at- full scale and left the basket is reported once, after the search, as- `insufficient_capital`, which the rebalance path's dropped sleeves now use- too. `zero_size` means the lot rule and nothing else.+- **A real VaR result grounds the VaR it returned** (#1464). `quantlib_call`records a scalar result under the function name. For `historical_var` and`parametric_var` that name reduces to `var`, which the grounding gate acceptsonly as a whole field name, so the result recorded no evidence and a correctanswer quoting it was rejected. Both names are now tail-risk aliases.Variances such as `residual_var` stay unmapped.-- **Twenty alphas no longer fill a missing input with a constant** (#1463).- A halt left a comparison reading 0 or False, a `.where(cond, 0)` reading a- flat day, or an `np.fmax` returning the other side, and the value then fed- every rolling window that reached back to the gap. Those cells are NaN now.- On gap-free data every finite value is bit-identical to before. A- perturbation sweep over all 462 alphas counts 35 still carrying a missing- bar forward, down from 55, and every one of them is accounted for: 28 are- recursive statistics (the GTJA `SMA(A, n, m)` smoothers and one running- product), which by the policy set on #1463 skip a missing observation and- continue from their last state — the operator layer's header now states- that exception, and a test pins it; 3 declare their own partial window; 4- are the sweep's own rank-tie artifact. Nothing rewarms a smoother after a- gap, and the registry's look-back mask was measured and rejected: it hid- 4,270 legitimate cells to remove 41% of the fabricated ones.-- **A `local:` code in a backtest is served from your own dataset or not at- all** (#1467). The market-data tool and the README already treated- `local:AAPL.US` this way, but the backtest runner did not. With- `source="local"` it counted the served `AAPL.US` as missing and sent it down- the A-share network fallback chain. With `source="auto"` it ignored the- prefix and fetched the symbol from network loaders. The prefix now picks the- local loader only, a symbol the dataset lacks fails with- `incomplete data for source=local`, and the rest of the run sees the bare- symbol: market rules, artifacts (`ohlcv_AAPL.US.csv`, no colon in the file- name) and the run card. A `local:` code requested from a network source, or- one symbol requested both with and without the prefix, is refused up front.-- **Event-study z-statistics no longer over-reject** (#1466). A CAR's standard- error summed each day's variance, ignoring that every day is forecast with- the same estimated parameters, so the days' abnormal returns are correlated.- On null data, the standardised CAR's variance was 1.11 with a 120-day- estimation window and 1.41–1.46 with a 30-day one. It is now 1.02 and 1.07,- which is only the t correction for an estimated residual variance. Patell and- BMP inherit the fix; `market_adjusted` was already right.-- **Robinhood live trading reads and trades the account the mandate names**- (#1442). The runner, the pre-trade gate and the commit-time ceiling fetch- called Robinhood with no `account_number`, and read replies one level too- shallow (`data.positions` where Robinhood nests `data.data.positions`). So- every reconciliation aborted, and the gate refused every order because it- could not read positions. The tests stayed green only because their fake- replies had a shape Robinhood never sends. A Robinhood mandate is now bound- at commit to one account from `get_accounts`: listed, not deactivated, and- open to agentic trading. The Web confirm dialog and the CLI both ask for- it, and never take it from the agent-written proposal. Every runner and gate- call sends that account, and an order naming another account is refused.- Held positions, which Robinhood reports without a price, are priced the way- a quantity order already was, and one that cannot be priced fails the- exposure check closed. Existing Robinhood mandates must be committed again- with an account.- **Plain integers in prose are not price claims.** A list number ("输出原则4"), a window length or a count was checked like a quoted price and could becut from a released answer.+- **A drawdown against an observed high passes wherever the endpoints arewritten.** "较 5 月高点 1.053 元已回撤约 37%" was rejected when its endpoints saton another line; a `derived` declaration does not depend on layout.+- **Price-denominated indicator values are evidence, by registration.**Moving averages, Bollinger bands and similar levels returned by`technical_indicators` are registered per tool as observed price evidence;RSI, MACD and other non-price leaves are not, and nothing is decided from afield name's words.+- **A year-less date is a date, and "24.6M" is 24.6 million.** Two real runsthat followed the contract were still released redacted: "2026-09-11 /09-14" in a table cell shipped as "2026-09-11 /(略※)-(略※)", and@@ -538,6 +578,7 @@ This project adheres to [Semantic Versioning](https://semver.org/)./ B / 千 / 万 / 亿 glued to a figure scale its comparison with the evidence,never its shape, and a cut removes the mark with the figure. Replayed onboth runs' tool results, each second draft now passes as written.+- **Metadata counts are not metrics, and a tool's tail-risk result can grounda VaR** (#1420, #1426). `return_observations`, `n_returns`, `return_window`,`vol_lookback`, `max_drawdown_duration`, `aligned_days` and their `*_window` /@@ -547,51 +588,880 @@ This project adheres to [Semantic Versioning](https://semver.org/).evidence, so a correct VaR is no longer refused. A field's metric comes fromthe head of its name, so `sharpe_sample_size` and `drawdown_threshold` areno longer a Sharpe ratio or a drawdown.+- **A correction names the prints of the figure's own instrument** (#1433). The"nearest observed" values come from that symbol (and its table column or`ref`), not from every field of every symbol in the run.+- **A number is read the way the chat renders it** (#1418). Zero-width- characters, a full-width or Arabic decimal point, a no-break space before- "%", the U+2212 minus sign, a backslash escape ("0\\.888"), an HTML entity- and an ISO code glued to the digits ("CNY0.888") no longer split a price into- unchecked integers; a decimal comma ("0,666 CNY", "12,5 %") is a decimal- while a valid grouping ("-1,250.00") stays grouped, and an answer that writes- a marked decimal comma reads "2,237" and "−5,132%" beside it as decimals too. Fences follow CommonMark,- and an unterminated code fence no longer exempts the rest of the answer.+ characters, a full-width or Arabic decimal point, a no-break space before "%",+ the U+2212 minus sign, a backslash escape ("0\\.888"), an HTML entity and an+ ISO code glued to the digits ("CNY0.888") no longer split a price into+ unchecked integers; a decimal comma ("0,666 CNY", "12,5 %") is a decimal while+ a valid grouping ("-1,250.00") stays grouped, and an answer that writes a+ marked decimal comma reads "2,237" and "−5,132%" beside it as decimals too.+ Fences follow CommonMark, and an unterminated code fence no longer exempts the+ rest of the answer.+- **A year is not a price shield.** "$2050", "1999 元" and a 2031 under a closecolumn are checked; a price placed in a date or code column is checked; acompact date, a dotted date and a time are structure. 円 and 원 are currencymarks, and "52pp" / "5200bp" are measurements.+- **A declaration cannot relocate or launder a figure.** A figure the sentencewrites about one instrument cannot be declared another's close; acurrency-marked result cannot be derived from an RSI or a volume; `count`,`derived` and `proposed` cannot sit in a price column.+- **A general answer with no tool evidence is cut, not refused.** A questionabout no instrument ("印花税怎么收") whose figures the gate cannot check isreleased with those figures omitted and a footnote, instead of a refusalabout prices it never mentioned; a market answer that observed no pricestill falls back. A malformed figures block no longer forces the fallbackeither: the block is dropped and the draft is checked as written.+- **The tool-call-syntax fallback reaches the chat once.** When a model answeredthe forced-text iteration with tool-call markup in a run whose output isbuffered, the replacement message was streamed by its own branch and again asthe released answer, so the chat and the CLI showed it twice.+- **No unchecked number reaches the chat before the gate runs.** An unbufferedanswer (no instrument asked about, no tool evidence yet) streamed its firstdraft live, numbers included, and a rejection then replaced it; the stream nowstops at the first measurement-shaped number, and a rejected draft's shownprefix is cleared with `stream_reset`.+- **A trading plan in a bare code fence is checked.** Only a fence tagged with alanguage holds code; an untagged fence is read as prose.+- **An integer price of an instrument quoted in the thousands is checked**("600519.SH 最新收盘 1520"), within half to twice its observed range; awindow such as "200 日均线" stays unchecked.+- **A price declared `count` is checked.** A number inside its instrument'sobserved price range may be a `count` only when a declared derivation uses itas a factor.+- **A redaction footnote the model wrote itself is removed** before the realcuts, so a released answer never carries two contradictory footnotes or amarker for a cut that did not happen.++- **A figure about a second instrument is checked against that instrument's+ evidence** (#1584, thanks @zeus229) — symbol resolution saw only+ price-comparable records, so a line naming an instrument the session held+ only fundamentals for took the answer's primary symbol and failed as+ `not_in_referenced_call`. Any observed evidence with an explicit symbol now+ resolves it; `47a036d6` scopes analysis metrics likewise, so a metric from a+ call about another symbol, or several, cannot ground it.++- **A field ref may contain `|`, and a correction for an ambiguous field lists+ the exact refs to use** (#1586, thanks @zeus229) — the `figures` parser+ split on every `|`, cutting such a ref short; a declaration is four columns+ now, the last taking the rest of the line. When several calls returned a+ field, the correction lists each valid `call_id::field` (only the figure's+ own symbol's since `47a036d6`), and a `tool_name::field` ref is refused as+ `field_ref_needs_call_id` with that list.++- **A decimal-comma answer's dot-grouped integers are read as thousands**+ (#1588, thanks @zeus229) — `AR$5.165; volumen 502.408` was read as two+ decimals and failed as `figure_undeclared`. Once a document writes a decimal+ comma, now including a `0,` with two or more digits, `5.165` is 5165 and is+ still checked; without that evidence it stays 5.165. The PR's wider 1%+ rounding band was not kept (`47a036d6`): at 0.5%, 0.82467 has to be written+ `0,825`, not `0,82`.++- **Figures from `technical_indicators` keep the date the tool reported**+ (#1590, thanks @zeus229) — the tool dates its snapshot as `latest_date`, a+ key evidence ingestion did not read, so `latest_close`, the indicators and+ the volume metrics landed in `grounding_evidence.json` with+ `timestamp: null`. It is inherited now, below a row's own date fields and+ above `as_of`; an undated payload stays undated.++#### Market data++- **`get_market_data` asks for `AAPL.US` instead of returning empty frames for+ a bare US ticker** — only codes the market detector reads as US equities+ are refused; a Chinese futures code (`RB0`, `IF2412`) or a joined crypto+ pair (`BTCUSDT`) is served as written. The public-source health canary+ probes Yahoo with `AAPL.US` for the same reason.++- **A mixed-market backtest falls back per market** — a symbol missing from+ its primary source was retried along the fallback chain of the first+ symbol's market, so a Hong Kong gap in an A-share-led universe was never+ retried on a Hong Kong source. Each missing symbol now follows its own+ market's chain, and the incomplete-data error lists which symbols were+ served.++- **A dollar-quoted BYMA or TSX line, or a Hong Kong RMB or USD counter, no+ longer enters a single-currency backtest** (#1576, after #1543, thanks+ @Shizoqua, @zeus229). Both venues list USD lines beside their home-currency+ ones. On 2026-09-24 Yahoo declared USD for GGALD.BA (4.20), SPYD.BA and+ AAPLD.BA, and for DLR-U.TO and XUS-U.TO. `ar_equity` and `ca_equity` are each+ one static-currency pool, so such a line was priced as pesos or Canadian+ dollars. As on the LSE, the loaders now read the declared currency. They admit+ a `.BA` line only in ARS and a `.TO` / `.V` line only in CAD. A trailing `D`+ does not decide it: YPFD.BA is a peso line. Every other declared quote+ currency is recorded in market-data provenance, and the grounding gate asks an+ answer to name that currency rather than the one the suffix implies. Hong Kong+ is decided by the code, because the Hong Kong chain starts with sources that+ declare no currency: HKEX's Stock Code Allocation Plan assigns 80000-89999 to+ products traded in renminbi and several 09xxx ranges (plus 10900-10999 and+ 41500-41599) to products traded in USD. So an RMB counter (80700.HK, 375.40+ CNY on 2026-09-24) or a USD ETF (9834.HK) is refused in a Hong Kong or+ cross-market backtest, is counted in its own currency by the mixed-currency+ guard, and is quoted in that currency by the grounding gate, whichever source+ served it. The table matched the currency Yahoo declared for all 24 codes+ probed across the ranges.++- **Grounding identity and symbol search know every market the data layer+ routes** (#1565, #1575, thanks @Shizoqua). `.BA`, `.L` and `.VN` symbols had no+ canonical-symbol scan, venue or currency in grounding. Symbol search+ labelled Indian, Korean, British, Vietnamese and Argentine results `global`.+ A parity test now takes the market list from the backtest's own currency+ table. An answer may write £, ₫ or AR$ for the currency it has to name.++- **A `local:` code in a backtest is served from your own dataset or not at+ all** (#1467). The market-data tool and the README already treated+ `local:AAPL.US` this way, but the backtest runner did not. With+ `source="local"` it counted the served `AAPL.US` as missing and sent it down+ the A-share network fallback chain. With `source="auto"` it ignored the+ prefix and fetched the symbol from network loaders. The prefix now picks the+ local loader only, a symbol the dataset lacks fails with+ `incomplete data for source=local`, and the rest of the run sees the bare+ symbol: market rules, artifacts (`ohlcv_AAPL.US.csv`, no colon in the file+ name) and the run card. A `local:` code requested from a network source, or+ one symbol requested both with and without the prefix, is refused up front.++- Block trades, margin trading and financial statements no longer read+ Eastmoney rejecting a stale query (code 9501) as an empty result; all six+ datacenter callers share `eastmoney_client.datacenter_rejection`.++- Lockup expiry: `free_shares` is the unlocking quantity+ (`CURRENT_FREE_SHARES`), not the float; units are stated (万股 / 万元) (#1513,+ #1501, thanks @lorenzozanee). The report is `RPT_LIFT_STAGE`; the old one+ rejected six columns.++- Dragon-tiger seats come from the buy/sell detail reports, surface a provider+ rejection, and rank within each listing reason (#1512, #1502, thanks+ @lorenzozanee).++- Shareholder-count history returns real periods from `RPT_HOLDERNUM_DET`,+ states each row's previous period, and warns when the average-holding join+ fails (#1518, #1503, thanks @he-yufeng).++- QVeris resolves adjusted fields as a complete set, one family per response,+ and no longer gives an unadjusted bar the adjusted volume (#1527, #1494,+ thanks @cgycorey).++- **Tencent history walks backward from the end date** (#1411, closes #1410,+ thanks @shadowinlife) — `fqkline` serves the *last* 500 bars of a window,+ not the first: `sh000300` over 2018-01-01..2026-06-30 comes back as 500 bars+ starting 2024-06-06 (verified live). The forward walk that shipped in 0.1.15+ (#1154) therefore returned only the last 500 bars of any longer window, with+ exit code 0. The cursor now moves back from `end_date`, and an error-shaped+ or mid-walk empty reply is retried instead of being read as the start of+ history.++- **A BaoStock server that stops answering no longer hangs the fetch** (#1615,+ closes #1492, thanks @he-yufeng) — baostock 0.9.3 reads its socket with no+ timeout, so a silent server blocked forever and a closed connection spun at+ 100% CPU. The connection, each complete message and the wait for the session+ are now bounded by `VIBE_TRADING_BAOSTOCK_TIMEOUT` (30 seconds by default),+ and a failed call is a per-symbol miss the chain moves past. After review,+ BaoStock sessions also run one at a time per process, because the SDK keeps+ one global socket, and a socket that failed mid-reply is closed so the next+ query cannot read its late answer.++- **Concurrent first lookups no longer find an empty loader registry** (#1405,+ closes #1403, thanks @alanwilhelm) — `_ensure_registered` marked itself done+ before importing the loader modules, so a thread arriving during a cold+ start returned with nothing registered and an available symbol came back+ `_unresolved`. The flag is published after the imports now, under a lock,+ and concurrent cold callers wait for them.++- **An explicit source that must not fall back no longer has its gaps filled**+ (#1441, thanks @chiww) — `local`, `fmp`, `qveris`, `tickerall`, `nobitex`+ and `wallex` were kept from degrading only when the loader was unavailable+ as a whole. A symbol the named source lacked was still fetched down the+ market's chain, so one run could mix a local snapshot with Tencent bars+ behind two log lines. The run now fails naming the missing symbols, pinned+ by a test for every source in that set.++- **The Stooq challenge warning names a variable that exists** (#1402, thanks+ @chiww, thanks @chiww) — it told operators to set+ `VIBE_TRADING_MARKET_DATA_ORDER_*`, which nothing reads, or to drop stooq from+ the chain, which override validation refuses. It names `MARKET_DATA_ORDER_*`+ now (for example `MARKET_DATA_ORDER_US_EQUITY`), suggests moving stooq to the+ end, and says an override must be a permutation of the default chain.++- **QVeris no longer sends the end date as `indicators`** (#1496, thanks+ @lorenzozanee, thanks @lorenzozanee) — end-date parameters were recognised by+ substring, so any capability parameter containing `to`, such as `indicators`,+ was given the end date. Only exact names are mapped now (`end`, `end_date`,+ `enddate`, `to`, `to_date`, `until`, `until_date`, with hyphens folded). The+ rest of #1494 was fixed separately.++- **`get_shareholder_count` answers again** (#1490, closes #1489, thanks+ @cgycorey, thanks @cgycorey) — it asked `RPT_HOLDERNUMLATEST` for+ `AVG_HOLD_AMT`, a column the report no longer has, so Eastmoney rejected the+ whole request and the tool reported "no shareholder-count disclosure found"+ for every A-share. It requests `AVG_MARKET_CAP`, the same per-account value+ (the output key `avg_hold_amount` is unchanged), and a rejected request now+ returns the datacenter's own message instead of an empty result.++- **`get_market_data` no longer answers `1M` with one-minute bars** (#1487,+ closes #1480, thanks @shadowinlife) — intervals were matched through an+ upper-case fold under which `1m` (a minute) and `1M` (a month) share a key,+ so a monthly request was served minute bars without an error. An exact-case+ match now wins before any fold. The PR refused `1M` and `1W`; since #1479,+ in this same release, they return monthly and weekly bars built from daily+ ones.++#### Backtest runner and engines++- **Backtests align calendars at any timestamp resolution** (#1560, thanks+ @Shizoqua). The engine merged symbol calendars and placed prices through raw+ nanosecond integers, but a local duckdb source serves microseconds. Such a run+ was dated 1970-01-21, and beside a nanosecond source one symbol's closes came+ back empty. The rebalance mask had the same unit mismatch.++- **Mean-variance style optimizers size a short by its own expected return**+ (#1548, thanks @Shizoqua). The optimizers scored every position as if it were+ held long, so the strongest shorts (most negative drift) looked worst and got+ the least capital. `BaseOptimizer.optimize` now negates a short's column+ before building the context, so `mu` is the position's expected return AND+ `cov` is the position covariance `D Σ D`: a long and a short of two +0.92+ correlated names hedge instead of reading as correlated (the old objective put+ the whole book on the long, the new one splits it 0.514 / -0.486). Risk-parity+ and max-diversification contexts get the position covariance with it;+ volatility-only contexts are unchanged, since std(-r) == std(r).++- **Monte-Carlo validation annualises at the venue's bars per year** (#1546,+ thanks @Shizoqua). It hardcoded 252 while `bootstrap_sharpe_ci` and+ `walk_forward_analysis` in the same report used the resolved value, so a+ crypto (365) or forex (260) run published two disagreeing Sharpes.++- **A futures order on a negative-price bar is no longer rejected outright**+ (#1547, thanks @Shizoqua). `FuturesBaseEngine` dropped the `abs(price)` its+ base class relies on under `allow_nonpositive_prices`, so size came out+ negative and the caller's `size <= 0` guard refused every order.++- **A backtest aborted when a funding debit left cash below zero** (#1542).+ CompositeEngine and CryptoEngine subtract crypto funding from capital with+ no floor, and the next open then fitted at no scale, not even an empty plan:+ the open-basket search kept the full-scale plan and the run died with+ "planned order … exceeds available capital", and a rebalance raised+ "insufficient capital for position rebalance". Such an open is now skipped+ and reported once as `insufficient_capital`, the bar's reductions still+ run, and a close whose loss exceeds its margin still aborts the bar (#1274).++- **One cash-starved open is one rejection, not twenty-six** (#1470). In+ `position_adjustment="hold"`, the search that scales a basket down to the+ cash available re-planned every open on each step, and every step whose+ size rounded to zero booked a `zero_size` rejection for the same symbol on+ the same bar — one contract the cash could not hold read as 26 lot-rounding+ failures. The trial plans are silent now; a sleeve that was a real order at+ full scale and left the basket is reported once, after the search, as+ `insufficient_capital`, which the rebalance path's dropped sleeves now use+ too. `zero_size` means the lot rule and nothing else.++- **A-shares from a source without its own engine branch run under A-share+ rules** (#1431, thanks @chiww) — for such sources the runner followed the+ instrument's market only for US, HK, Canadian and UK equities, so+ `600519.SH` served by `local`, `tencent`, `eastmoney`, `baostock`, `mootdx`+ or `sina` ran on `CryptoEngine`: no stamp tax, T+1, price limits or+ 100-share lots, and a perpetual funding fee charged against the position.+ They route to `ChinaAEngine` now.++- **The declared interval is checked against the bars that arrived** (#1432,+ thanks @chiww) — a single-market run annualised at the interval it asked+ for, but a loader can serve coarser bars (the local loader cannot turn a+ daily file into `1H`), and CAGR, Sharpe and volatility then scale by the+ ratio. The runner compares the median bar spacing of the longest price frame+ with the declared interval; past a 1.5x gap either way it annualises as the+ interval the spacing matches, or from the calendar when bars four or more+ days apart match none, and the report is written to the run card beside the+ caliber warning. Median spacing replaced a bars-per-calendar-year count,+ which measures five daily bars starting on a Monday at 456 a year.++- **Options annual return is annualised over the elapsed bars** (#1458, thanks+ @Shizoqua, thanks @Shizoqua) — the options engine raised growth to+ `bars_per_year / (n - 1)` where the shared `calc_metrics` uses `bars_per_year+ / n`, which exaggerated annual return in both directions, most on short runs.+ Both use `n` now.++- **An India short is covered against the buy-side circuit band** (#1608,+ thanks @Shizoqua) — the India engine never passed the held position's+ direction to the shared band check, so covering a short, which is a buy, was+ tested against the lower band a long's sale uses, and a cover at a locked+ upper circuit filled. The direction is passed now, in single-market and+ composite runs alike.++#### The backtest tool and what a run leaves behind++- **The backtest tool honours `VIBE_TRADING_TOOL_TIMEOUT_SECONDS` and keeps a+ timed-out run's output** — the engine subprocess had a hardcoded 300-second+ limit regardless of the setting (default 1,800; 0 disables it), and a+ timeout discarded everything the engine had printed. The partial stdout and+ stderr are now written to the run's `logs/` and the last 2,000 characters+ returned with the error.++- **Runs from the `backtest` tool record `state.json`** (#1413, closes #1412,+ thanks @shadowinlife) — only the agent runtime wrote it and+ `refresh_strategy_evidence` refuses a run without one, so every run from the+ `backtest` tool, or the MCP tool that delegates to it, was rejected as run+ status unknown despite complete artifacts. The tool now records success,+ failure with the exit code, or a timeout, through the same `RunStateStore`+ writer; a config that fails validation records nothing, since the engine+ never ran.++- **The `backtest` tool names the directory it looked in** (#1429, thanks+ @cgycorey, thanks @cgycorey) — a missing `config.json` or+ `code/signal_engine.py` was reported with no path, so a caller whose `run_dir`+ resolved somewhere unexpected could not tell that from an empty directory.+ Both errors now name the resolved directory and say where the file belongs.++- **Dict- and list-valued metrics reach the run card** (#1424, thanks+ @cgycorey, thanks @cgycorey) — the card kept only scalar metrics, so+ `unfilled_plan_rejections_by_symbol`, which names the rejected sleeve and+ the reason, was dropped while the anonymous count beside it stayed.+ Non-scalar metrics now go to a `structured_metrics` block, shown in the Run+ Card tab and the Markdown card, with `metrics` unchanged; a value over 4 KiB+ of card JSON, or one that cannot be serialised, is left out and named under+ `_omitted`.++- **A text column in `metrics.csv` no longer drops a run's metrics from the+ reply** (#1579, thanks @Shizoqua) — the session service converted the whole+ row at once and swallowed the error, so a run with an explicit benchmark,+ whose `benchmark_ticker` is text, lost every metric without a log line. Each+ field is converted on its own now, and only the unconvertible one is+ skipped.++#### Alpha Zoo++- **24 alphas emitted a value computed from a missing bar** (#1463, #1523,+ #1534, #1452, thanks @Shizoqua). A comparison over an operand that is NaN+ because its window holds the gap is False, and the #1463 sweep, which nudged+ the bar by a tick, could not see that. A second oracle compares the gapped run+ with the gap-free one. Comparison gates and `np.fmin`/`np.fmax` now mask on+ their inputs' reach (`factors.base.observed_over`) rather than on their+ operands' NaN, so a correlation undefined on complete data keeps its old+ verdict; the six fmax/fmin alphas converted on 2026-09-18 had lost up to 87%+ of their post-warmup values in a 5-symbol universe that way and are restored.+ On gap-free data no finite value changed.++- **Twenty alphas no longer fill a missing input with a constant** (#1463).+ A halt left a comparison reading 0 or False, a `.where(cond, 0)` reading a+ flat day, or an `np.fmax` returning the other side, and the value then fed+ every rolling window that reached back to the gap. Those cells are NaN now.+ On gap-free data every finite value is bit-identical to before. A+ perturbation sweep over all 462 alphas counts 35 still carrying a missing+ bar forward, down from 55, and every one of them is accounted for: 28 are+ recursive statistics (the GTJA `SMA(A, n, m)` smoothers and one running+ product), which by the policy set on #1463 skip a missing observation and+ continue from their last state — the operator layer's header now states+ that exception, and a test pins it; 3 declare their own partial window; 4+ are the sweep's own rank-tie artifact. Nothing rewarms a smoother after a+ gap, and the registry's look-back mask was measured and rejected: it hid+ 4,270 legitimate cells to remove 41% of the fabricated ones.++- **Asset growth is year over year on the daily panel** (#1564, thanks+ @Shizoqua). Derived fundamentals run on a panel densified to the price+ calendar, where a one-row lag compared yesterday with today: 0% on ordinary+ days and a one-day spike at each filing. The lag is now 365 days, as-of.++- **A factor's IC ratio is unset for a negative baseline** (#1549, thanks+ @Shizoqua). Dividing two negative IC means gives a positive ratio, so a+ rolling IC that got further below zero reported as improvement. The signal+ itself took the worst of the available metrics and was not fooled; the+ published number was.++- **qlib158 `CNTP` / `CNTN` / `CNTD` stop counting a missing close as a flat+ day** (#1459, thanks @cgycorey) — the 15 factors (three families × 5 / 10 /+ 20 / 30 / 60 windows) turn `close > prev close` into 0/1 and average it,+ and a NaN comparison reads `False`. The registry mask from #1377 blanks the+ gap bar itself, but every later window that still held the gap counted the+ halt as a day that neither rose nor fell, so `cntd` read closer to zero+ than the observed days say. The comparison is NaN now wherever either close+ is missing: a gap blanks `window + 1` rows and the value returns, equal to+ the gap-free run, once it scrolls out. The same rule moves the first value+ from row `window - 1` to row `window`, matching each factor's+ `min_warmup_bars`, because the first bar has no prior close. On sparse+ panels this can trip the registry's >95% NaN guard; each factor's `notes`+ states from what missing-bar fraction.++- **qlib158 `WVMA` takes the absolute return in its numerator, as upstream+ does** (#1594, thanks @Shizoqua) — `wvma5` … `wvma60` computed+ `ts_std(ret·v) / ts_mean(|ret|·v)`, a signed numerator over an absolute+ denominator, where qlib's formula takes `Abs` on both. Every value of the+ five factors changes. The finishing commit (`a5d47c20`) points the+ attribution header at the upstream file that actually defines WVMA+ (`qlib/contrib/data/loader.py`, MIT) instead of `handler.py`, and pins the+ formula against an independent scalar oracle, a gap and all-zero volume.++- **GTJA alphas 103, 133 and 177 score a fresh extreme high, not low**+ (#1604, thanks @Shizoqua) — `HIGHDAY` / `LOWDAY` count bars since the+ extreme, but the port put the 0-based window index, counted from the oldest+ bar, in their place. Alpha 103 read a low set today as 5 and a low 19 bars+ old as 100: the recency signal was inverted. The PR restored the direction+ one bar off (95 and 0). The finishing commit (`0244ecea`) counts days since+ the extreme with today as 0, so alphas 103 and 177 run from 5 to 100 over+ a 20-bar window, and takes the most recent bar on a tie. It also corrects+ `gtja191/LICENSE.md` to cite the Guotai Junan report of 2017-06-15 by Li+ Chen and Liu Fubing, where the zoo had named a 2014 report.++- **A custom zoo root no longer replaces the bundled alpha of the same name**+ (#1468, fixes #1465, thanks @cgycorey) — `Registry(zoo_root=...)` loaded+ custom alpha files into `sys.modules` under the bundled module's own name.+ Once a custom root had computed its own `qlib158/cntp5.py`, a default+ `Registry()` in the same process computed the custom code too. The import+ cache is keyed by module path and zoo root now; module names and relative+ imports inside a custom root are unchanged. No shipped caller passes+ `zoo_root`; it is the path tests and plugins use.++#### Quant Library and analysis tools++- **A strict-bench OOS split needs two IC observations on each side** (#1559,+ thanks @Shizoqua). A split outside the loaded prices is refused. An alpha+ whose IC series leaves fewer than two observations on either side is skipped,+ with the counts in the reason. The IC series ends one forward-return horizon+ before the prices, so a split one bar from the end used to publish verdicts+ measured on nothing.++- **Quant inputs are checked before they are used** (#1555-#1558, thanks+ @Shizoqua). HRP aligns a supplied correlation matrix, and the covariance's own+ rows, to the covariance's column labels; swapping two rows used to move the+ weights with no error. Purged cross-validation refuses an unordered or+ duplicated timestamp index, and a label that ends before it starts. Impact+ models and every fixed-income entry point refuse NaN and infinities instead of+ returning NaN prices, durations and curves.++- **Shadow-account overtrading uses the whole trading window** (#1563, thanks+ @Shizoqua). A long hold that closed before later short trades no longer+ shrinks the span.++- **Event-study z-statistics no longer over-reject** (#1466). A CAR's standard+ error summed each day's variance, ignoring that every day is forecast with+ the same estimated parameters, so the days' abnormal returns are correlated.+ On null data, the standardised CAR's variance was 1.11 with a 120-day+ estimation window and 1.41–1.46 with a 30-day one. It is now 1.02 and 1.07,+ which is only the t correction for an estimated residual variance. Patell and+ BMP inherit the fix; `market_adjusted` was already right.++- Shadow account: an `other`-market rule no longer matches every symbol (#1538,+ thanks @Shizoqua), and overtrading PnL no longer double-counts trades already+ explained (#1475).++- Quantlib: comps refuses negative bridge magnitudes like the DCF bridge (#1535,+ thanks @Shizoqua); accrued interest is exact on coupon dates under 30/360 and+ rejects an unknown day count there too (#1509); Kyle's lambda fits an+ intercept (#1477).++- ML strategy: a single-class training window is skipped and the previous model+ keeps predicting (#1522, thanks @Shizoqua). Memory: two entries with one title+ and different types keep separate index rows, and `forget` names the type when+ ambiguous (#1525). Swarm run and task files use unique temp names (#1536).++- **Patell z uses each normal-return model's own degrees of freedom** (#1449,+ thanks @Shizoqua) — the `df / (df - 2)` correction in `event_study`+ assumed the market model's two estimated parameters for every event, so+ `market_adjusted` (none estimated) and `mean_adjusted` (one) over-stated+ the variance and reported a slightly smaller |z|, and a larger p-value,+ than their own fits support. The market model is unchanged. The finishing+ commit (`844b13b9`) has `estimate_market_model` read its residual degrees+ of freedom from the same table, so the two cannot drift apart.++- **Microstructure estimators refuse NaN and infinite inputs** (#1451,+ thanks @Shizoqua) — a non-finite price came back from+ `roll_effective_spread` as a `0.0` spread, `kyles_lambda` returned `nan`,+ and `vpin` produced a series from NaN or negative volumes. All three raise+ now, and `vpin` also refuses a non-finite `bucket_size`.+ `amihud_illiquidity` drops an infinite row the way it already dropped a NaN+ one; before, an infinite return made the ratio infinite and an infinite+ volume counted as a zero-illiquidity day. The accept side is pinned too+ (`c6010208`): constant prices are a zero spread, all-zero order flow a zero+ lambda, and zero-volume bars leave VPIN unchanged.++- **Copula pseudo-observations give ties one rank and leave NaN missing**+ (#1453, thanks @Shizoqua) — the double `argsort` ranked tied values by+ position (`[1, 1, 3]` became 0.25 / 0.5 / 0.75) and sorted a NaN last,+ handing it the column's top rank and counting it in the normaliser.+ `pseudo_observations` uses average ranks now, keeps a NaN as NaN with each+ column normalised by its own count of observations, and refuses infinite+ values.++- **`spread_term_structure` keeps tenor `5` and `5.0` in one column** (#1507,+ thanks @Shizoqua) — the column name was built from the raw key, so issuers+ quoting the same tenor as an int and as a float were split across+ `5Y_spread_bp` and `5.0Y_spread_bp`, each NaN for the other group. The+ reference-curve lookup already treated the two as one tenor.++- **Purged cross-validation refuses a timestamp label index it cannot map**+ (#1556, thanks @Shizoqua) — label end times passed as a Series are turned+ into positions with `searchsorted`, which assumes a sorted index. An+ unordered or duplicated index mapped label ends to the wrong rows with no+ error, so the purge protected the wrong rows; it raises now. The finishing+ commit (`250ef746`) closes the neighbouring gap: a timestamp label ending+ before its own start was clipped up to it (ends at positions `[1, 0, 3, 3]`+ came back as `[1, 1, 3, 3]`) and now raises, as the positional form already+ did. A `NaT` end stays allowed and purges through the last observation.++- **Market-impact models refuse infinite inputs** (#1557, thanks @Shizoqua)+ — NaN was already refused, but positive infinity passed every positivity+ check: an infinite ADV priced a `linear_impact` or `sqrt_impact` order at+ zero impact, and an infinite price, traded volume or `bps` returned an+ infinite buy fill. Price, volume, ADV, `bps`, impact coefficient,+ volatility and `eta` must all be finite now.++- **The correlation-analysis cointegration examples call the Quant Library**+ (#1404, thanks @Shizoqua) — the skill's Engle-Granger example carried its+ own statsmodels fit and raised `KeyError` on an unnamed Series, because it+ looked the hedge ratio up by the series' name. The PR named the regressor;+ the finishing commit (`bd29a187`) replaced the local fit with+ `find_hedge_ratio`, `cointegration_test` and `adf_test` from+ `src.quantlib.timeseries`, and the half-life example with+ `compute_half_life`. The example now inherits their refusal to pair two+ legs on different indices positionally, keeps its result keys and adds+ `half_life`.++- **`portfolio_risk_xray` works on windows longer than 250 bars** (#1462,+ closes #1461, thanks @he-yufeng) — the tool fetched prices under the shared+ 250-row cap. Above it `fetch_market_data` returns a sampled envelope rather+ than the bar list the tool parses, so a window of 251 bars or more failed+ with "no close prices returned for any requested symbol", which is about+ anything over a year of daily bars. It requests every row now+ (`max_rows=0`, as `technical_indicators` does); volatility, VaR/ES and+ drawdown need consecutive bars regardless.++- **`benford_check` reads the leading digit from the number as written**+ (#1591, thanks @Shizoqua) — `financial_rigor`'s Benford check took the+ digit from `10 ** frac(log10(v))`, which lands just under an integer for+ some exact round values, so `5000` counted as a leading 4. Among the+ integers 1 to 100,000, twelve were misread (8, 30, 40, 50, 80, 500, 600,+ 900, 5000, 6000, 9000, 50000). The PR rounded before truncating, which+ would have read `4999.9999999` as a 5; the finishing commit (`d0ef98a6`)+ takes the first significant digit of a `Decimal` built from the supplied+ value instead, with no float round-trip. Zero, NaN and infinite values are+ still skipped.++- **The sentiment scorer reads "low" as bearish only** (#1448, thanks @Shizoqua,+ thanks @Shizoqua) — "low" was in both the bullish and the bearish word lists,+ so it cancelled itself and `shares tumbled to a new low today` scored neutral.+ It was the only word in both; a test now keeps the lists disjoint.++- **Head-and-shoulders detection checks shoulder symmetry on negative+ series** (#1456, thanks @Shizoqua) — the `pattern` tool divided the+ shoulder gap by the signed shoulder average. On a series below zero the+ ratio went negative and always passed the 5% test, so clearly asymmetric+ shoulders were flagged as a pattern. It divides by the absolute average+ now, as `double_top_bottom` already did.++#### Live trading, brokers and the strategy store++- **Robinhood live trading reads and trades the account the mandate names**+ (#1442). The runner, the pre-trade gate and the commit-time ceiling fetch+ called Robinhood with no `account_number`, and read replies one level too+ shallow (`data.positions` where Robinhood nests `data.data.positions`). So+ every reconciliation aborted, and the gate refused every order because it+ could not read positions. The tests stayed green only because their fake+ replies had a shape Robinhood never sends. A Robinhood mandate is now bound+ at commit to one account from `get_accounts`: listed, not deactivated, and+ open to agentic trading. The Web confirm dialog and the CLI both ask for+ it, and never take it from the agent-written proposal. Every runner and gate+ call sends that account, and an order naming another account is refused.+ Held positions, which Robinhood reports without a price, are priced the way+ a quantity order already was, and one that cannot be priced fails the+ exposure check closed. Existing Robinhood mandates must be committed again+ with an account.++- **SDM decay metrics treat NaN and infinite bench results as missing**+ (#1447, thanks @0xouzm) — three stored `sharpe=inf` rows read as `healthy`,+ so a non-dry `sdm_decay_scan` moved a strategy in monitoring back to active,+ and `-inf` read as critical. SQLite keeps infinities as written, so the+ persistent store was affected too. `compute_decay_metrics` now drops+ non-finite values, independently for IC and Sharpe, before the+ three-observation minimum.++- **A mandate narrowed to zero exposure is stored as zero** (#1593, thanks+ @Shizoqua, thanks @Shizoqua) — narrowing `max_total_exposure_usd` or+ `account_funding_usd` to `0` at commit is legal, but the cap mapping used `x+ or default`, so the saved mandate got the order cap back as its exposure+ limit: the commit widened what the user had just narrowed. An explicit zero+ now persists, and an order checked against that mandate breaches+ `max_total_exposure_usd`.++- **A malformed Robinhood `buying_power` is a shape error, not an unknown**+ (#1526, thanks @lorenzozanee) — a non-null `buying_power` that was not an+ object (a bare number, string or list) was read as unknown and passed; it+ now raises, so the portfolio source errors instead of carrying the reply.+ An omitted or null value stays unknown and zero stays valid. As merged the+ PR also refused an omitted `buying_power`; `46892136` restored that case to+ unknown, which is what the PR description specified.++- **Interval and event triggers carry `market=None` again** (#1599, thanks+ @woshi77777stars, thanks @woshi77777stars) — `Trigger.market` was both a+ dataclass field and a classmethod factory, so every non-market trigger carried+ the bound method as its `market`, and `dataclasses.asdict()` + `json.dumps()`+ raised on it. `due_now()` was unaffected. The factory is bound after the class+ body (`d0ef98a6` keeps subclass binding intact), and a repo-wide static test+ fails on any dataclass or NamedTuple field shadowed this way.++- **With an MT5 profile selected, `search_symbol` resolves the broker's own+ symbols** (#1597, closes #1589, thanks @lorenzozanee) — it returned no MT5+ candidate for an explicit symbol such as `XAUUSD`, so the identity gate+ blocked `trading_quote`, `trading_history` and `get_market_data`. Search+ now reads the configured terminal's catalog by `terminal_path` with no login+ credentials. After `47a036d6` every query is answered from that terminal+ while MT5 is selected, a web-search hit is never substituted, and an+ ambiguous alias is refused.++- **Backtests attach to the running MT5 terminal without seeing its+ credentials** (#1598, thanks @he-yufeng) — the sandbox's ephemeral HOME had+ no `mt5.json`, so the loader attached bare and a forex backtest could fall+ through to akshare while `get_market_data` used MT5 in-process. The sandbox+ now gets only `terminal_path` and `timeout` (validated by `47a036d6`), never+ `login`, `password` or `server`, and a bare attach warns that the SDK may+ pick another installation logged into another account.++- **Dhan paper orders reject fractional share quantities** (#1595, thanks+ @Shizoqua, thanks @Shizoqua) — `0.5` passed the positive check and truncated+ to `0`, returning a successful simulated fill for zero shares, and `1.5`+ silently became `1`. Only whole numbers of at least one share are accepted;+ `a5d47c20` parses with `Decimal`, so `1.00000000000000001` is refused too, and+ NaN, infinity or text return an error instead of raising.++#### Swarm workers++- **A swarm worker can load only the skills its spec lists** (#1414, thanks+ @shadowinlife, thanks @shadowinlife) — `SwarmAgentSpec.skills` was read only+ to filter the prompt, so `load_skill` loaded any skill a worker named. The+ worker's `load_skill` now refuses a skill outside the list and names the ones+ allowed. An empty list stays unrestricted, as the prompt filter treats it; the+ main agent is unchanged.++- **A swarm tool that declares `run_dir` gets the directory the model named,+ confined to the agent's workspace** (#1430, thanks @cgycorey) — the worker+ overwrote every call's `run_dir` with the workspace, so `backtest`, exposed+ to 19 preset agents, never saw the directory it was given. A relative path+ now resolves under the workspace, an absolute one inside it is kept, and one+ outside is refused with an error naming the workspace. Tools that do not+ declare `run_dir` still get the workspace.++- **Every swarm worker that writes its summary also writes its message log**+ (#1439, thanks @cgycorey) — `messages.json` is the only record of the+ arguments the model asked a tool for, and the completed, token-limit,+ content-filter and incomplete exits wrote `summary.md` without it, so the+ most common outcome left no log. Both files now go through one writer. A+ failed LLM call still writes neither.++#### Agent loop, providers and goals++- **An explicit provider header survives an ambient twin** (#1573, #1568, thanks+ @zeus229). When `OPENAI_CUSTOM_HEADERS` names a header the provider also sets+ (a `user-agent` beside the Kimi / NVIDIA / OpenCode `User-Agent`), openai+ 3.19.2's case-insensitive merge dropped the provider's value. That turned CI+ red for every PR on 2026-09-24. The first fix proposed would have sent both+ headers under openai 2.53, the version the lock file and the Docker image+ install. Every ambient spelling is now omitted and the explicit value re-set+ after the omits, which holds under both.++- **Read-only results lost to context compaction are restored, not refetched**+ (#1488, thanks @zeus229). A successful read-only call whose payload compaction+ removed is replayed from the run's own cache, at most six times a run; past+ the cap it runs again, as it did before, instead of being refused with "use+ the previous result" for a payload the model cannot see. The cap is checked+ per restore, and any successful write empties the cache, so a regenerated file+ is read afresh. `read_url` opts in; `no_cache=true` always fetches.++- **OpenAI's gpt-5.6 models run with their reasoning on the agent's default+ configuration** (#1473). `/v1/chat/completions` refuses function tools for+ `gpt-5.6-terra` and `gpt-5.6-luna` unless `reasoning_effort` is `none`,+ including when no effort is configured and the model applies its own+ default, so a plain `LANGCHAIN_PROVIDER=openai` setup failed on its first+ tool call with `Function tools with reasoning_effort are not supported`.+ The adapter now treats that refusal the way it treats a rejected+ `temperature` or `stream_options`: the request is retried on+ `/v1/responses`, where the effort travels as `reasoning.effort`, and the+ model is remembered for the rest of the process so later calls go there+ first. Nothing changes for a model whose chat endpoint accepts tools, and+ `LANGCHAIN_USE_RESPONSES_API=true` still selects the route up front and+ skips the one failed request.++- **A research goal that is an order is still refused** (#1562, thanks+ @Shizoqua). The execution filter no longer rejects research that mentions+ shares or coins ("should the fund sell its GOOGL shares"). It refuses an+ objective that opens with buy or sell, puts a quantity right after the verb,+ or writes 买入 / 卖出 with a quantity and a unit. The first version of the change+ had made "Buy 100 shares of NVDA" an accepted goal.++- **Research goals refuse an order that puts "now" or "immediately" before+ the verb** (#1605, thanks @Shizoqua) — the execution filter matched only+ the verb-then-cue order, so `Immediately buy TSLA`, `Now sell GOOGL` and+ `Please immediately short TSLA` were accepted as research objectives. The+ mirrored rule requires the cue to sit directly before the verb, with only+ separators and an optional `please` between them. The first version allowed+ up to 40 characters there and also refused `Now research whether to buy+ AAPL`; the finishing commit (`0244ecea`) keeps that and similar research+ phrasings accepted. The check runs on a goal's objective, on each of its+ criteria and on an objective update.++- **The tail a full compaction keeps counts a thinking model's reasoning**+ (#1582, thanks @Shizoqua) — the layer-3 tail is sized to a 20K-token budget,+ but the sizing read `content` and tool-call arguments and skipped+ `reasoning_content`, so on reasoning-heavy turns the kept tail could run far+ past its budget and compaction freed less than intended. The reasoning text+ is counted now, as the loop's token estimate already counted it.++- **`technical_indicators` results are restored after compaction instead of+ fetched again** (#1601, thanks @zeus229) — the tool is read-only but+ `repeatable`, and a repeatable tool is replayed from the run's memory only if+ it opts in; it did not, so an identical request after compaction re-ran the+ market-data fetch. It opts in now, like `web_reader`, within the loop's+ existing replay budget.++- **Token counts and Polymarket CLOB token ids are no longer redacted as+ credentials** (#1606, thanks @Shizoqua) — the `token` credential marker is a+ substring match, so `token_budget`, `token_usage`, `total_input_tokens`,+ `total_output_tokens` and `clob_token_id` read `[redacted]` in traces, event+ previews and the live audit ledger. Since `0244ecea` each is released only in+ its expected shape (a non-negative integer, a map of counts, a decimal+ uint256 id); any other value under those names stays redacted.++- **A remote MCP tool with a malformed `properties` schema still gets a valid+ one** (#1607, thanks @Shizoqua) — `normalize_mcp_tool_schema` dropped a+ non-object `properties` such as a string or a list and then skipped the `{}`+ fallback, so the tool reached the model as an object schema with no+ `properties` key at all. It gets `{}` now, unless the top level composes+ `anyOf` / `oneOf` / `allOf`.++- **The Copilot credential fallback re-reads the token after 60 seconds**+ (#1619, closes #1617, thanks @lorenzozanee) — when neither+ `COPILOT_GITHUB_TOKEN` nor `OPENAI_API_KEY` is set, `get_llm_credentials`+ falls back to `GH_TOKEN` / `GITHUB_TOKEN` or `gh auth token`, and it cached+ that answer for the life of the process. The cache now expires after 60+ seconds. Its consumer is provider diagnostics: Copilot requests go through+ the SDK client, which resolves credentials each time it is created.++#### Channels, memory and configuration++- **The Email channel verifies TLS certificates by default** (#1544, thanks+ @shadowinlife). The implicit-SSL paths (`IMAP4_SSL` / `SMTP_SSL`) in both the+ connection probe and the polling/send adapter used Python's default unverified+ context (`CERT_NONE`, no hostname check), so the mailbox password travelled to+ a server whose certificate was never checked — an active network attacker+ between the operator and their mail provider could harvest it. All four call+ sites now build their context through a shared `email_tls_context()` that+ verifies the certificate and hostname against the system CA bundle, matching+ the STARTTLS path, which already did. A new `verify_tls` field (default+ `true`, surfaced in the guided Email setup in all nine locales) is the+ documented opt-out for self-signed or internal-CA servers. **Upgrade note:** a+ mail server with such a certificate connected before this change and now fails+ the TLS handshake until `verify_tls` is set to `false` for it.++- **Plain IMAP no longer sends the mailbox password in clear text** (after+ #1544). With `imap_use_ssl` off, the Email channel and its connection test+ logged in over an unencrypted connection, and the guided setup put that+ switch one click away. A new `imap_use_tls` (on by default, the IMAP twin of+ `smtp_use_tls`) upgrades the connection with STARTTLS before `LOGIN`; a+ server that offers no STARTTLS is reported as a network failure before the+ password is sent. `verify_tls` now governs STARTTLS on IMAP and SMTP as well+ as implicit SSL, so a self-signed server can opt out on any path. **Upgrade+ note:** an `imap_use_ssl: false` setup against a server without STARTTLS+ stops connecting until `imap_use_tls` is set to `false` for it, which sends+ the password in plain text again.++- **Settings-write routes reject cross-site browser requests** (#1544, thanks+ @shadowinlife). `require_settings_write_auth` did not apply the cross-site+ guard that its siblings `require_auth` and `require_event_stream_auth` enforce+ on unsafe methods, so a malicious web page could attempt a CSRF write against+ any settings route — including the new `POST /channels/email/test`, which+ merges stored credentials with a caller-supplied patch and could be steered to+ send the stored mailbox password to an attacker-chosen host. The guard now+ runs first on every settings-write route; same-origin Web UI calls and+ non-browser clients (CLI/curl, no `Origin`) are unaffected.++- Channel adapters log through stdlib again: `{}` placeholders, napcat's 17+ sites and loguru-only `.opt()` calls lost their messages (signal's+ `_safe_handle` raised instead of swallowing) (#1533, #1531, thanks+ @shadowinlife). Slack DMs from unapproved senders get a pairing code (#1524,+ thanks @Shizoqua); Discord no longer pings `@everyone` or roles (#1537);+ duplicate-reply suppression covers integer message ids and is bounded (#1476);+ the WebSocket channel starts again (from #1521).++- **Two memories sharing a title keep separate index rows and links** (#1545,+ thanks @Shizoqua). Under `VT_MEMORY_HIERARCHY` an entry lives at+ `{memory_type}/{slug}.md`, so two entries with one title had the same+ `path.name`: the second `add()` overwrote the first's `MEMORY.md` row, and the+ semantic-link block beside it excluded the other entry as "self" and wrote an+ ambiguous target. Both key on the path relative to the memory dir now; a+ sidecar written earlier, with a bare filename or an absolute path, is still+ read.++- `/agent.json`, `/agent.yaml` and `/agent.yml` are gitignored at the runtime+ root; the Feishu QR login already wrote channel credentials there.++- **Signal channel: stripping the bot's mention no longer removes the wrong+ character after an emoji** (#1478, thanks @Shizoqua) — Signal reports a+ mention's `start` and `length` in UTF-16 code units, and the channel sliced+ the Python string with them directly. Each character outside the Basic+ Multilingual Plane (most emoji) before the mention shifted the slice by+ one, so the mention placeholder stayed in the text handed to the agent and+ the neighbouring character was cut instead. Offsets are converted to string+ indices before a mention is stripped or tested as a leading mention.++- **A file received over DingTalk is saved inside its download directory+ whatever it is named** (#1506, thanks @Shizoqua) — the adapter joined the+ inbound `fileName` to the per-sender download path unsanitized, so a name+ with `../` segments, or an absolute path (which replaces the whole path when+ joined), could write outside it. The name goes through `safe_filename` now,+ as the other adapters already sanitize a remote file name.++- **A channel that fails to load logs which one and why** (#1580, thanks+ @Shizoqua, thanks @Shizoqua) — two channel-registry log calls had two `%s`+ placeholders and one argument, so an external plugin that failed to import, or+ a built-in channel skipped for a missing SDK (at debug level), printed `---+ Logging error ---` instead of its name and the exception.++- **The README says which process owns Telegram polling** (#1583, closes+ #1553, thanks @lorenzozanee) — Telegram allows one `getUpdates` poller per+ bot token. `vibe-trading serve` owns the channel runtime, and+ `channels start` asks the running server (`VIBE_TRADING_API_URL`, default+ `http://127.0.0.1:8000`) to start it rather than polling itself, so the+ conflict to avoid is a second server process with the same token enabled.++- **Memory garbage collection no longer compresses an entry it just archived**+ (#1450, thanks @Shizoqua) — the compression pass reused the entry list read+ before archiving, so an entry just moved to `archive/` was compressed from a+ path that no longer existed and logged a failure. `451c554a` skips by whether+ the file still exists, not by id: ids are six hex characters two files can+ share, and an archive step that failed leaves its file to be compressed.++- **An unparseable optional numeric setting falls back to its default instead+ of failing config load** (#1592, thanks @Shizoqua) — the coercion that drops+ a bad `int` / `float` env value skipped `int | None` fields, so+ `ANTHROPIC_MAX_TOKENS=not_a_number` raised while building `EnvConfig`.+ Optional fields get the same fallback now; a parseable value outside the+ field's bounds (`ANTHROPIC_MAX_TOKENS=0`) is still rejected.++#### Web UI and locales++- **Leaving a chat cancels its pending history scrolls** — a scroll scheduled+ while a conversation's history loaded could still fire after leaving it,+ and a history load still in flight could schedule another.++- **A welcome-screen quick action stays highlighted after you click it**+ (#1500, thanks @shadowinlife) — the first chip was styled as selected by its+ position, whatever had been clicked, so the chip you chose never lit up. The+ highlight now follows the click and moves with the next one, and each chip+ reports `aria-pressed`.++- **Broker SDK diagnostics no longer name Longbridge for every broker**+ (#1436, thanks @tingkk) — the shared `sdk_missing` and `network_unreachable`+ messages on the Runtime page and in the chat connector status said+ Longbridge whichever broker raised them, so an unreachable Futu OpenD gateway+ sent the user to check Longbridge. They now name the affected broker's SDK,+ service or local gateway, in every locale.+## [0.1.15] — 2026-09-09Rolls up 551 commits / 162 merged pull requests since 0.1.14, from 35diff --git a/Dockerfile b/Dockerfileindex bf04cea79..daa09423e 100644--- a/Dockerfile+++ b/Dockerfile@@ -61,7 +61,7 @@ FROM python:3.11-slim@sha256:e031123e3d85762b141ad1cbc56452ba69c6e722ebf2f042cc0LABEL org.opencontainers.image.title="Vibe-Trading" \org.opencontainers.image.description="Natural-language finance research AI agent with backtesting" \- org.opencontainers.image.version="0.1.15" \+ org.opencontainers.image.version="0.1.16" \org.opencontainers.image.source="https://github.com/HKUDS/Vibe-Trading" \org.opencontainers.image.licenses="MIT"diff --git a/README.md b/README.mdindex a51b90de7..8c578314a 100644--- a/README.md+++ b/README.md@@ -52,20 +52,22 @@> ⚠️ **Security warning:** The X account `VibeTrading_HKU`, Virtuals project `101845`, and token contract `0x640BDBF77b6447E8b7DB7894cED84BD1c40571f4` are not official Vibe-Trading assets. We have never launched or endorsed any token or memecoin. Do not buy, connect a wallet, or sign anything. [Details](SECURITY.md#official-channels--impersonation).-- **2026-09-29** 🛠️ **Long comparisons finish, and backtest reports keep their numbers**: Conversation compaction is now sized to the model's real context window (272K for GPT-6 on Codex, 1M for DeepSeek V4) instead of a fixed 40K estimate that began discarding tool results from the third step, so a 12-company comparison re-fetched the same filings until the no-progress stop. Failed runs keep their steps and the stop message names the tool and the reason; on Codex, mid-run instructions no longer replace the system prompt, reasoning is carried between turns, and the default model is `gpt-6-sol`. The figure check now reads a backtest's own output (Sortino, Calmar, turnover, weights, Monte Carlo p-values) and accepts a figure that names the run it came from, so a risk-parity vs equal-weight report is no longer replaced by a refusal. `get_market_data` asks for `AAPL.US` instead of returning empty data for a bare US ticker, and a mixed-market backtest falls back per market.+- **2026-09-29** 🚀 **v0.1.16 released** ([Release notes](https://github.com/HKUDS/Vibe-Trading/releases/tag/v0.1.16), `pip install -U vibe-trading-ai`): 492 commits and 116 merged pull requests since 0.1.15, from 16 contributors. **The theme of this cycle is a number that can show where it came from.** The grounding gate no longer guesses what a number is from the words around it: the model declares each figure's role (observed, derived, proposed, cited or count), the gate checks it against the session's tool evidence, and a figure that fails is cut instead of the whole answer being refused. A backtest's own output — Sortino, turnover, weights and Monte Carlo p-values, not only Sharpe — now grounds the report written about it, and run cards cite a metric only where the saved CSV agrees. Long research runs keep what they fetched: compaction follows each model's real context window instead of a fixed 40K estimate. The missing-data sweep reached 59 more alphas. **Fixed in this release, from a user's report:** failed runs keep their steps and the stop message names the call that produced nothing; `read_file` says where it may read instead of letting the model guess paths until the no-progress stop; on Codex, mid-run instructions no longer replace the system prompt, reasoning carries between turns and the default model is `gpt-6-sol`; and `get_market_data` asks for `AAPL.US` rather than returning empty data for a bare US ticker. **New:** Argentina (BYMA), weekly and monthly bars, Bahasa Indonesia, Gildata, OpenCode, Email and WebSocket channels, IM channel setup in the Web UI, and KIS, Upbit, Toss Securities and Scalable Capital connectors, 18 brokers in all. Thanks [@Shizoqua](https://github.com/Shizoqua), [@zeus229](https://github.com/zeus229), [@cgycorey](https://github.com/cgycorey), [@shadowinlife](https://github.com/shadowinlife), [@lorenzozanee](https://github.com/lorenzozanee), [@he-yufeng](https://github.com/he-yufeng), [@chiww](https://github.com/chiww), [@as950118](https://github.com/as950118), [@woshi77777stars](https://github.com/woshi77777stars), [@Yoruxyv](https://github.com/Yoruxyv), [@sambazhu](https://github.com/sambazhu), [@0xouzm](https://github.com/0xouzm), [@tingkk](https://github.com/tingkk), [@tonydo](https://github.com/tonydo), [@alanwilhelm](https://github.com/alanwilhelm) and [@modelpath-dev](https://github.com/modelpath-dev)!- **2026-09-28** 🛠️ **Source health and connector transparency**: scheduled public-source canaries report outages and data drift with bounded, credential-free probes ([#1627](https://github.com/HKUDS/Vibe-Trading/pull/1627)); BaoStock socket requests now have deadlines and serialize concurrent sessions ([#1615](https://github.com/HKUDS/Vibe-Trading/pull/1615)). Copilot credential lookups refresh after a short cache TTL ([#1619](https://github.com/HKUDS/Vibe-Trading/pull/1619)). A generated broker matrix preserves each profile’s paper/live permissions and labels declared capabilities separately from runtime verification ([#1629](https://github.com/HKUDS/Vibe-Trading/pull/1629)). **Grounding and MT5 follow-up:** explicit evidence symbols and exact field references are preserved; localized grouped figures retain the existing numeric tolerance ([#1584](https://github.com/HKUDS/Vibe-Trading/pull/1584), [#1586](https://github.com/HKUDS/Vibe-Trading/pull/1586), [#1588](https://github.com/HKUDS/Vibe-Trading/pull/1588)). Text-only correction attempts are bounded ([#1600](https://github.com/HKUDS/Vibe-Trading/pull/1600)). MT5 search uses the selected terminal and refuses ambiguous broker aliases; backtest sandboxes receive only validated attachment settings ([#1597](https://github.com/HKUDS/Vibe-Trading/pull/1597), [#1598](https://github.com/HKUDS/Vibe-Trading/pull/1598)).- **2026-09-27** 🛠️ **Verifiable run cards and safer research**: Run cards now show hashed backtest execution records and verified metric-to-CSV references in JSON, Markdown and Run Detail ([#1612](https://github.com/HKUDS/Vibe-Trading/pull/1612)). GTJA high/low recency factors use the correct day count and consistent tie handling ([#1604](https://github.com/HKUDS/Vibe-Trading/pull/1604)); India short covers check the buy-side circuit band ([#1608](https://github.com/HKUDS/Vibe-Trading/pull/1608)). Token usage remains readable without exempting arbitrary credential strings ([#1606](https://github.com/HKUDS/Vibe-Trading/pull/1606)), malformed MCP schemas retain their object properties ([#1607](https://github.com/HKUDS/Vibe-Trading/pull/1607)), and immediate-order goal checks preserve research questions ([#1605](https://github.com/HKUDS/Vibe-Trading/pull/1605)). Setup documentation now correctly identifies OpenRouter as the shipped default ([#1609](https://github.com/HKUDS/Vibe-Trading/issues/1609)).+<details>+<summary>Earlier news</summary>+- **2026-09-26** 🛠️ **More reliable indicators and trading limits**: Technical indicators retain their observation date and can reuse results lost during conversation compaction ([#1590](https://github.com/HKUDS/Vibe-Trading/pull/1590), [#1601](https://github.com/HKUDS/Vibe-Trading/pull/1601)). All five Qlib158 WVMA windows now use absolute returns in the numerator ([#1594](https://github.com/HKUDS/Vibe-Trading/pull/1594)); Benford checks preserve the correct leading digit at numeric boundaries ([#1591](https://github.com/HKUDS/Vibe-Trading/pull/1591)). An explicitly zero exposure limit stays zero ([#1593](https://github.com/HKUDS/Vibe-Trading/pull/1593)), and Dhan paper orders reject fractional or invalid quantities without rounding away input precision ([#1595](https://github.com/HKUDS/Vibe-Trading/pull/1595)). Invalid optional numeric environment variables fall back to defaults while explicit configuration constraints remain enforced ([#1592](https://github.com/HKUDS/Vibe-Trading/pull/1592)). Runtime triggers keep the correct field defaults and subclass factory behavior ([#1599](https://github.com/HKUDS/Vibe-Trading/pull/1599)).- **2026-09-25** 🛠️ **More reliable research sessions and channel diagnostics**: A text column in a backtest CSV no longer discards the numeric metrics ([#1579](https://github.com/HKUDS/Vibe-Trading/pull/1579)), and conversation compaction now counts reasoning content in its retained-message budget ([#1582](https://github.com/HKUDS/Vibe-Trading/pull/1582)). Failed channel loads now log the actual exception ([#1580](https://github.com/HKUDS/Vibe-Trading/pull/1580)); the Telegram guide clarifies that CLI and Web controls share the same API runtime ([#1583](https://github.com/HKUDS/Vibe-Trading/pull/1583)). Robinhood portfolio reads reject malformed buying-power objects while keeping omitted or null values unknown ([#1526](https://github.com/HKUDS/Vibe-Trading/pull/1526)). Leaving a chat also cancels pending history-scroll timers. Thanks [@Shizoqua](https://github.com/Shizoqua) and [@lorenzozanee](https://github.com/lorenzozanee)!-- **2026-09-24** 💱 **A dollar line priced as pesos, Email and WebSocket in the Web UI, and a CI break from an SDK update**: BYMA and the TSX list US-dollar lines beside their home-currency ones (GGALD.BA, DLR-U.TO), and an Argentine or Canadian backtest priced them in its single peso or Canadian-dollar pool. The loaders now admit a line only in its market's currency, as LSE lines already were, and every other declared quote currency travels with the data, so an answer names the right one ([#1576](https://github.com/HKUDS/Vibe-Trading/pull/1576)). Hong Kong's RMB and USD counters (80700.HK, 9834.HK), which HKEX numbers in their own code ranges, are refused in a Hong Kong backtest for the same reason and quoted in their own currency. Stock profiles carry the listing's own venue and currency beside the issuer's fundamentals ([#1577](https://github.com/HKUDS/Vibe-Trading/pull/1577)), grounding and symbol search know every market the data layer routes ([#1575](https://github.com/HKUDS/Vibe-Trading/pull/1575)), and `technical_indicators` reports volume with the unit its source declares — board lots or shares, 100× apart ([#1571](https://github.com/HKUDS/Vibe-Trading/pull/1571)). Email and WebSocket join the guided channel setup, with mail-server certificates now verified and settings writes closed to cross-site pages, and a plain IMAP login is upgraded with STARTTLS before the password goes out ([#1544](https://github.com/HKUDS/Vibe-Trading/pull/1544)). **Fixed:** an explicit provider header lost to an ambient twin under openai 3.19.2, which had turned every PR's CI red ([#1568](https://github.com/HKUDS/Vibe-Trading/pull/1568)); a backtest from a microsecond-resolution source was dated 1970 ([#1560](https://github.com/HKUDS/Vibe-Trading/pull/1560)); a research goal that is plainly an order is refused again ([#1562](https://github.com/HKUDS/Vibe-Trading/pull/1562)); a strict-bench OOS split could leave one side empty ([#1559](https://github.com/HKUDS/Vibe-Trading/pull/1559)); HRP label alignment, cross-validation label order and non-finite quant inputs ([#1555](https://github.com/HKUDS/Vibe-Trading/pull/1555), [#1556](https://github.com/HKUDS/Vibe-Trading/pull/1556), [#1557](https://github.com/HKUDS/Vibe-Trading/pull/1557), [#1558](https://github.com/HKUDS/Vibe-Trading/pull/1558)); the shadow-account overtrading window ([#1563](https://github.com/HKUDS/Vibe-Trading/pull/1563)); and asset growth was read day over day ([#1564](https://github.com/HKUDS/Vibe-Trading/pull/1564)). Thanks [@Shizoqua](https://github.com/Shizoqua), [@zeus229](https://github.com/zeus229) and [@shadowinlife](https://github.com/shadowinlife)!-<details>-<summary>Earlier news</summary>++- **2026-09-24** 💱 **A dollar line priced as pesos, Email and WebSocket in the Web UI, and a CI break from an SDK update**: BYMA and the TSX list US-dollar lines beside their home-currency ones (GGALD.BA, DLR-U.TO), and an Argentine or Canadian backtest priced them in its single peso or Canadian-dollar pool. The loaders now admit a line only in its market's currency, as LSE lines already were, and every other declared quote currency travels with the data, so an answer names the right one ([#1576](https://github.com/HKUDS/Vibe-Trading/pull/1576)). Hong Kong's RMB and USD counters (80700.HK, 9834.HK), which HKEX numbers in their own code ranges, are refused in a Hong Kong backtest for the same reason and quoted in their own currency. Stock profiles carry the listing's own venue and currency beside the issuer's fundamentals ([#1577](https://github.com/HKUDS/Vibe-Trading/pull/1577)), grounding and symbol search know every market the data layer routes ([#1575](https://github.com/HKUDS/Vibe-Trading/pull/1575)), and `technical_indicators` reports volume with the unit its source declares — board lots or shares, 100× apart ([#1571](https://github.com/HKUDS/Vibe-Trading/pull/1571)). Email and WebSocket join the guided channel setup, with mail-server certificates now verified and settings writes closed to cross-site pages, and a plain IMAP login is upgraded with STARTTLS before the password goes out ([#1544](https://github.com/HKUDS/Vibe-Trading/pull/1544)). **Fixed:** an explicit provider header lost to an ambient twin under openai 3.19.2, which had turned every PR's CI red ([#1568](https://github.com/HKUDS/Vibe-Trading/pull/1568)); a backtest from a microsecond-resolution source was dated 1970 ([#1560](https://github.com/HKUDS/Vibe-Trading/pull/1560)); a research goal that is plainly an order is refused again ([#1562](https://github.com/HKUDS/Vibe-Trading/pull/1562)); a strict-bench OOS split could leave one side empty ([#1559](https://github.com/HKUDS/Vibe-Trading/pull/1559)); HRP label alignment, cross-validation label order and non-finite quant inputs ([#1555](https://github.com/HKUDS/Vibe-Trading/pull/1555), [#1556](https://github.com/HKUDS/Vibe-Trading/pull/1556), [#1557](https://github.com/HKUDS/Vibe-Trading/pull/1557), [#1558](https://github.com/HKUDS/Vibe-Trading/pull/1558)); the shadow-account overtrading window ([#1563](https://github.com/HKUDS/Vibe-Trading/pull/1563)); and asset growth was read day over day ([#1564](https://github.com/HKUDS/Vibe-Trading/pull/1564)). Thanks [@Shizoqua](https://github.com/Shizoqua), [@zeus229](https://github.com/zeus229) and [@shadowinlife](https://github.com/shadowinlife)!- **2026-09-23** 🇦🇷 **Argentine listings, a short book sized upside down, and the last half of the tail-risk gate**: `.BA` symbols — BYMA equities and CEDEARs — route to Yahoo as their own ARS market, and the Web UI groups and labels them like any other; an Argentine backtest still fails closed until BYMA execution rules are modelled ([#1543](https://github.com/HKUDS/Vibe-Trading/pull/1543)). The mean-variance and turnover-aware optimizers scored every position as if it were held long, so the strongest shorts got the least capital. Both now size on position returns, where a long and a short of two correlated names hedge each other instead of reading as correlated — on a +0.92 correlated pair the old objective put the whole book on the long, the new one splits it 0.51 / -0.49 ([#1548](https://github.com/HKUDS/Vibe-Trading/pull/1548)). And a tail-risk figure now has to name its own field once a session holds more than one measure or confidence level, so an ES 95% can no longer quote the VaR 95% value under a call id ([#1425](https://github.com/HKUDS/Vibe-Trading/issues/1425)). **Fixed:** Monte-Carlo validation annualises at the venue's bars per year instead of a hardcoded 252, so its Sharpe agrees with the rest of the same report ([#1546](https://github.com/HKUDS/Vibe-Trading/pull/1546)); a futures order on a negative-price bar is no longer rejected outright ([#1547](https://github.com/HKUDS/Vibe-Trading/pull/1547)); a factor's IC ratio is left unset for a negative baseline instead of reading real decay as healthy improvement ([#1549](https://github.com/HKUDS/Vibe-Trading/pull/1549)); and two memories sharing a title under different types keep their own index rows and link targets ([#1545](https://github.com/HKUDS/Vibe-Trading/pull/1545)). Thanks [@zeus229](https://github.com/zeus229), [@Shizoqua](https://github.com/Shizoqua) and [@he-yufeng](https://github.com/he-yufeng)!diff --git a/README_ar.md b/README_ar.mdindex 3f3816bdf..ee3eedbc9 100644--- a/README_ar.md+++ b/README_ar.md@@ -52,20 +52,22 @@> ⚠️ **تحذير أمني:** حساب X باسم `VibeTrading_HKU`، ومشروع Virtuals رقم `101845`، وعقد التوكن `0x640BDBF77b6447E8b7DB7894cED84BD1c40571f4` كلّها غير رسمية ولا تتبع Vibe-Trading. لم نُطلق أو نؤيد مطلقًا أي توكن أو عملة ميم. لا تشترِ هذا التوكن، ولا تربط محفظتك، ولا توقّع أي شيء. [التفاصيل](SECURITY.md#official-channels--impersonation).-- **2026-09-29** 🛠️ **المقارنات الطويلة تكتمل، وتقارير الاختبار الرجعي تحتفظ بأرقامها**: أصبح ضغط المحادثة يُحسب وفق نافذة السياق الفعلية للنموذج (272K لـ GPT-6 عبر Codex و1M لـ DeepSeek V4) بدلًا من تقدير ثابت قدره 40K كان يحذف نتائج الأدوات بدءًا من الخطوة الثالثة، فكانت مقارنة 12 شركة تعيد جلب القوائم المالية نفسها حتى التوقف بسبب عدم التقدم. تحتفظ عمليات التشغيل الفاشلة بخطواتها، وتذكر رسالة التوقف الأداة والسبب؛ وفي Codex لم تعد التعليمات المضافة أثناء التشغيل تستبدل موجّه النظام، ويُنقل الاستدلال بين الأدوار، والنموذج الافتراضي هو `gpt-6-sol`. يقرأ فحص الأرقام الآن مخرجات الاختبار الرجعي نفسه (Sortino وCalmar ومعدل الدوران والأوزان وقيم p لمونت كارلو) ويقبل الرقم الذي يسمّي التشغيل الذي جاء منه، فلم يعد تقرير مقارنة تكافؤ المخاطر بالأوزان المتساوية يُستبدل برسالة رفض. يطلب `get_market_data` الرمز `AAPL.US` بدلًا من إرجاع بيانات فارغة لرمز أمريكي بلا لاحقة، ويعود الاختبار الرجعي متعدد الأسواق إلى مصادر بديلة لكل سوق على حدة.+- **2026-09-29** 🚀 **صدور v0.1.16** ([ملاحظات الإصدار](https://github.com/HKUDS/Vibe-Trading/releases/tag/v0.1.16), `pip install -U vibe-trading-ai`): 492 عملية إيداع و116 طلب دمج منذ 0.1.15، من 16 مساهمًا. **موضوع هذه الدورة: رقمٌ يستطيع أن يبيّن من أين جاء.** لم تعد بوابة التحقق تخمّن ماهية الرقم من الكلمات المحيطة به: يعلن النموذج دور كل رقم (observed أو derived أو proposed أو cited أو count)، وتتحقق البوابة منه مقابل نتائج أدوات الجلسة، ويُحذف الرقم الذي لا يجتاز التحقق بدل رفض الإجابة كلها. أصبحت مخرجات الاختبار الرجعي نفسه — Sortino ومعدل الدوران والأوزان وقيم p لمونت كارلو، لا نسبة شارب وحدها — أساسًا للتحقق من التقرير المكتوب عنه، ولا تستشهد بطاقات التشغيل بمقياس إلا إذا طابق ملف CSV المحفوظ. لم تعد مهام البحث الطويلة تفقد ما جلبته: يتبع ضغط السياق نافذة السياق الفعلية لكل نموذج بدل تقدير ثابت قدره 40K. وامتدت مراجعة البيانات الناقصة إلى 59 عامل ألفا إضافيًا. **إصلاحات من بلاغ مستخدم:** تحتفظ عمليات التشغيل الفاشلة بخطواتها وتذكر رسالة التوقف الاستدعاء الذي لم يُنتج شيئًا؛ ويوضح `read_file` أين يمكنه القراءة بدل أن يظل النموذج يخمّن المسارات حتى التوقف لعدم التقدم؛ وفي Codex لم تعد التعليمات المضافة أثناء التشغيل تستبدل موجّه النظام، ويُنقل الاستدلال بين الأدوار، والنموذج الافتراضي هو `gpt-6-sol`؛ ويطلب `get_market_data` الرمز `AAPL.US` بدل إرجاع بيانات فارغة لرمز أمريكي بلا لاحقة. **الجديد:** الأرجنتين (BYMA)، والشموع الأسبوعية والشهرية، والإندونيسية، وGildata، وOpenCode، وقناتا البريد الإلكتروني وWebSocket، وإعداد قنوات المراسلة من واجهة الويب، وموصلات KIS وUpbit وToss Securities وScalable Capital، ليبلغ عدد الوسطاء 18. شكرًا [@Shizoqua](https://github.com/Shizoqua), [@zeus229](https://github.com/zeus229), [@cgycorey](https://github.com/cgycorey), [@shadowinlife](https://github.com/shadowinlife), [@lorenzozanee](https://github.com/lorenzozanee), [@he-yufeng](https://github.com/he-yufeng), [@chiww](https://github.com/chiww), [@as950118](https://github.com/as950118), [@woshi77777stars](https://github.com/woshi77777stars), [@Yoruxyv](https://github.com/Yoruxyv), [@sambazhu](https://github.com/sambazhu), [@0xouzm](https://github.com/0xouzm), [@tingkk](https://github.com/tingkk), [@tonydo](https://github.com/tonydo), [@alanwilhelm](https://github.com/alanwilhelm) و[@modelpath-dev](https://github.com/modelpath-dev)!- **2026-09-28** 🛠️ **سلامة مصادر البيانات ووضوح قدرات الوسطاء**: فحوص دورية للمصادر العامة دون بيانات اعتماد وبمهلة محددة ترصد أعطال الاتصال والبيانات ([#1627](https://github.com/HKUDS/Vibe-Trading/pull/1627)). أضيفت مهل لاتصالات BaoStock وتسلسل للجلسات المتزامنة ([#1615](https://github.com/HKUDS/Vibe-Trading/pull/1615))، وتجدد استعلامات بيانات اعتماد Copilot بعد انتهاء تخزين مؤقت قصير ([#1619](https://github.com/HKUDS/Vibe-Trading/pull/1619)). تفصل مصفوفة الوسطاء المولدة صلاحيات التداول التجريبي والفعلي لكل ملف، وتميز القدرات المعلنة عن التحقق التشغيلي ([#1629](https://github.com/HKUDS/Vibe-Trading/pull/1629)). **تحسينات التحقق وMT5**: تُحفظ رموز الأدوات الصريحة ومراجع الحقول الدقيقة، مع إبقاء حدود التسامح الرقمي عند قراءة الفواصل المحلية ([#1584](https://github.com/HKUDS/Vibe-Trading/pull/1584)، [#1586](https://github.com/HKUDS/Vibe-Trading/pull/1586)، [#1588](https://github.com/HKUDS/Vibe-Trading/pull/1588)). محاولات التصحيح النصي محدودة ([#1600](https://github.com/HKUDS/Vibe-Trading/pull/1600)). يبحث MT5 عبر المنصة المحددة ويرفض الأسماء البديلة الملتبسة؛ وتتلقى بيئة الاختبار المعزولة إعدادات اتصال متحققة فقط ([#1597](https://github.com/HKUDS/Vibe-Trading/pull/1597)، [#1598](https://github.com/HKUDS/Vibe-Trading/pull/1598)).- **2026-09-27** 🛠️ **بطاقات تشغيل قابلة للتحقق وتحسينات للبحث**: تعرض JSON وMarkdown وصفحة تفاصيل التشغيل سجلات تنفيذ الاختبار التاريخي بصماتٍ مشفّرة ومراجع موثقة لمقاييس CSV ([#1612](https://github.com/HKUDS/Vibe-Trading/pull/1612)). صُحّح حساب الأيام وحالات التعادل في عوامل GTJA ([#1604](https://github.com/HKUDS/Vibe-Trading/pull/1604))، ويُفحص إغلاق المراكز القصيرة في الهند وفق حد السعر لجهة الشراء ([#1608](https://github.com/HKUDS/Vibe-Trading/pull/1608)). تبقى أعداد استخدام الرموز مقروءة دون السماح بسلاسل اعتماد عشوائية ([#1606](https://github.com/HKUDS/Vibe-Trading/pull/1606))، وتُستعاد خصائص مخططات MCP غير الصالحة ([#1607](https://github.com/HKUDS/Vibe-Trading/pull/1607)). يسمح فحص أهداف الأوامر الفورية بأسئلة البحث ([#1605](https://github.com/HKUDS/Vibe-Trading/pull/1605))، وتوضح الوثائق أن OpenRouter هو الإعداد الافتراضي ([#1609](https://github.com/HKUDS/Vibe-Trading/issues/1609)).+<details>+<summary>أخبار سابقة</summary>+- **2026-09-26** 🛠️ **مؤشرات وحدود تداول أكثر موثوقية**: تحتفظ المؤشرات الفنية بتاريخ الرصد ويمكنها إعادة استخدام النتائج التي أزيلت أثناء ضغط المحادثة ([#1590](https://github.com/HKUDS/Vibe-Trading/pull/1590)، [#1601](https://github.com/HKUDS/Vibe-Trading/pull/1601)). تستخدم نوافذ WVMA الخمس في Qlib158 العوائد المطلقة في البسط ([#1594](https://github.com/HKUDS/Vibe-Trading/pull/1594))، ويحافظ فحص Benford على الرقم الأول الصحيح عند الحدود العددية ([#1591](https://github.com/HKUDS/Vibe-Trading/pull/1591)). يبقى حد التعرض المحدد صراحةً بصفر صفراً ([#1593](https://github.com/HKUDS/Vibe-Trading/pull/1593))، وترفض أوامر Dhan التجريبية الكميات الكسرية أو غير الصالحة دون فقدان دقة المدخلات بسبب التقريب ([#1595](https://github.com/HKUDS/Vibe-Trading/pull/1595)). تعود متغيرات البيئة العددية الاختيارية غير الصالحة إلى القيم الافتراضية مع استمرار تطبيق قيود الإعدادات الصريحة ([#1592](https://github.com/HKUDS/Vibe-Trading/pull/1592)). كما أُصلحت القيم الافتراضية لحقول مشغلات وقت التشغيل وسلوك إنشاء مثيلاتها في الأصناف الفرعية ([#1599](https://github.com/HKUDS/Vibe-Trading/pull/1599)).- **2026-09-25** 🛠️ **جلسات بحث وتشخيص قنوات أكثر موثوقية**: لم يعد وجود عمود نصي في ملف CSV للاختبار التاريخي يؤدي إلى فقدان المقاييس الرقمية ([#1579](https://github.com/HKUDS/Vibe-Trading/pull/1579))، وأصبح ضغط المحادثة يحتسب محتوى الاستدلال ضمن ميزانية الرسائل المحتفظ بها ([#1582](https://github.com/HKUDS/Vibe-Trading/pull/1582)). تُسجَّل الآن الاستثناءات الفعلية عند فشل تحميل القنوات ([#1580](https://github.com/HKUDS/Vibe-Trading/pull/1580))، ويوضح دليل Telegram أن أوامر CLI وعناصر تحكم الويب تشترك في بيئة تشغيل API نفسها ([#1583](https://github.com/HKUDS/Vibe-Trading/pull/1583)). ترفض قراءة محفظة Robinhood كائنات القوة الشرائية ذات البنية غير الصحيحة، مع إبقاء القيم المحذوفة أو null غير معروفة ([#1526](https://github.com/HKUDS/Vibe-Trading/pull/1526)). كما تُلغى مؤقتات تمرير السجل المعلقة عند مغادرة المحادثة. شكرًا [@Shizoqua](https://github.com/Shizoqua) و[@lorenzozanee](https://github.com/lorenzozanee)!-- **2026-09-24** 💱 **أوراق مسعّرة بالدولار حُسبت بالبيزو، وEmail وWebSocket في واجهة الويب، وعطل في CI سببه تحديث SDK**: تُدرج BYMA وبورصة تورنتو أوراقًا مسعّرة بالدولار الأمريكي إلى جانب أوراقها بالعملة المحلية (GGALD.BA وDLR-U.TO)، وكان الاختبار الرجعي الأرجنتيني أو الكندي يسعّرها في مجمّعه الوحيد بالبيزو أو بالدولار الكندي. صارت أدوات التحميل تقبل الورقة فقط إذا كانت مسعّرة بعملة سوقها، كما هو الحال مع أوراق LSE، وتنتقل عملة التسعير المعلنة لغيرها مع البيانات فيذكر الجواب العملة الصحيحة ([#1576](https://github.com/HKUDS/Vibe-Trading/pull/1576)). كما تُرفض في الاختبار الرجعي لهونغ كونغ، للسبب نفسه، الخطوط المتداولة باليوان والدولار (80700.HK و9834.HK) التي ترقّمها HKEX في نطاقات رموز خاصة بها، وتُعرض بعملتها. ملفات الأسهم تحمل بورصة الإدراج وعملته إلى جانب البيانات الأساسية للمُصدر ([#1577](https://github.com/HKUDS/Vibe-Trading/pull/1577))، ويعرف التحقق من الأرقام والبحث عن الرموز كل سوق تدعمه طبقة البيانات ([#1575](https://github.com/HKUDS/Vibe-Trading/pull/1575))، وتعيد `technical_indicators` حجم التداول مع الوحدة التي يعلنها المصدر — لوت أو سهم، والفرق 100 ضعف ([#1571](https://github.com/HKUDS/Vibe-Trading/pull/1571)). وينضم Email وWebSocket إلى الإعداد الموجَّه للقنوات، مع التحقق من شهادات خادم البريد ورفض كتابة الإعدادات من صفحات عابرة للمواقع، وترقية تسجيل دخول IMAP غير المشفّر عبر STARTTLS قبل إرسال كلمة المرور ([#1544](https://github.com/HKUDS/Vibe-Trading/pull/1544)). **إصلاحات:** ترويسة المزوّد الصريحة كانت تُستبدل بترويسة بيئية بالاسم نفسه في openai 3.19.2، ما جعل CI يفشل لكل PR ([#1568](https://github.com/HKUDS/Vibe-Trading/pull/1568))؛ واختبار رجعي من مصدر بدقة الميكروثانية كان مؤرخًا بعام 1970 ([#1560](https://github.com/HKUDS/Vibe-Trading/pull/1560))؛ وهدف بحثي هو أمر تداول صريح يُرفض مجددًا ([#1562](https://github.com/HKUDS/Vibe-Trading/pull/1562))؛ وتقسيم OOS في strict bench كان قد يترك أحد الجانبين فارغًا ([#1559](https://github.com/HKUDS/Vibe-Trading/pull/1559))؛ ومحاذاة تسميات HRP وترتيب تسميات التحقق المتقاطع والمدخلات غير المنتهية ([#1555](https://github.com/HKUDS/Vibe-Trading/pull/1555) و[#1556](https://github.com/HKUDS/Vibe-Trading/pull/1556) و[#1557](https://github.com/HKUDS/Vibe-Trading/pull/1557) و[#1558](https://github.com/HKUDS/Vibe-Trading/pull/1558))؛ ونافذة الإفراط في التداول في الحساب الظلي ([#1563](https://github.com/HKUDS/Vibe-Trading/pull/1563))؛ ونمو الأصول الذي كان يُحسب يومًا بيوم ([#1564](https://github.com/HKUDS/Vibe-Trading/pull/1564)). شكرًا [@Shizoqua](https://github.com/Shizoqua) و[@zeus229](https://github.com/zeus229) و[@shadowinlife](https://github.com/shadowinlife)!-<details>-<summary>أخبار سابقة</summary>++- **2026-09-24** 💱 **أوراق مسعّرة بالدولار حُسبت بالبيزو، وEmail وWebSocket في واجهة الويب، وعطل في CI سببه تحديث SDK**: تُدرج BYMA وبورصة تورنتو أوراقًا مسعّرة بالدولار الأمريكي إلى جانب أوراقها بالعملة المحلية (GGALD.BA وDLR-U.TO)، وكان الاختبار الرجعي الأرجنتيني أو الكندي يسعّرها في مجمّعه الوحيد بالبيزو أو بالدولار الكندي. صارت أدوات التحميل تقبل الورقة فقط إذا كانت مسعّرة بعملة سوقها، كما هو الحال مع أوراق LSE، وتنتقل عملة التسعير المعلنة لغيرها مع البيانات فيذكر الجواب العملة الصحيحة ([#1576](https://github.com/HKUDS/Vibe-Trading/pull/1576)). كما تُرفض في الاختبار الرجعي لهونغ كونغ، للسبب نفسه، الخطوط المتداولة باليوان والدولار (80700.HK و9834.HK) التي ترقّمها HKEX في نطاقات رموز خاصة بها، وتُعرض بعملتها. ملفات الأسهم تحمل بورصة الإدراج وعملته إلى جانب البيانات الأساسية للمُصدر ([#1577](https://github.com/HKUDS/Vibe-Trading/pull/1577))، ويعرف التحقق من الأرقام والبحث عن الرموز كل سوق تدعمه طبقة البيانات ([#1575](https://github.com/HKUDS/Vibe-Trading/pull/1575))، وتعيد `technical_indicators` حجم التداول مع الوحدة التي يعلنها المصدر — لوت أو سهم، والفرق 100 ضعف ([#1571](https://github.com/HKUDS/Vibe-Trading/pull/1571)). وينضم Email وWebSocket إلى الإعداد الموجَّه للقنوات، مع التحقق من شهادات خادم البريد ورفض كتابة الإعدادات من صفحات عابرة للمواقع، وترقية تسجيل دخول IMAP غير المشفّر عبر STARTTLS قبل إرسال كلمة المرور ([#1544](https://github.com/HKUDS/Vibe-Trading/pull/1544)). **إصلاحات:** ترويسة المزوّد الصريحة كانت تُستبدل بترويسة بيئية بالاسم نفسه في openai 3.19.2، ما جعل CI يفشل لكل PR ([#1568](https://github.com/HKUDS/Vibe-Trading/pull/1568))؛ واختبار رجعي من مصدر بدقة الميكروثانية كان مؤرخًا بعام 1970 ([#1560](https://github.com/HKUDS/Vibe-Trading/pull/1560))؛ وهدف بحثي هو أمر تداول صريح يُرفض مجددًا ([#1562](https://github.com/HKUDS/Vibe-Trading/pull/1562))؛ وتقسيم OOS في strict bench كان قد يترك أحد الجانبين فارغًا ([#1559](https://github.com/HKUDS/Vibe-Trading/pull/1559))؛ ومحاذاة تسميات HRP وترتيب تسميات التحقق المتقاطع والمدخلات غير المنتهية ([#1555](https://github.com/HKUDS/Vibe-Trading/pull/1555) و[#1556](https://github.com/HKUDS/Vibe-Trading/pull/1556) و[#1557](https://github.com/HKUDS/Vibe-Trading/pull/1557) و[#1558](https://github.com/HKUDS/Vibe-Trading/pull/1558))؛ ونافذة الإفراط في التداول في الحساب الظلي ([#1563](https://github.com/HKUDS/Vibe-Trading/pull/1563))؛ ونمو الأصول الذي كان يُحسب يومًا بيوم ([#1564](https://github.com/HKUDS/Vibe-Trading/pull/1564)). شكرًا [@Shizoqua](https://github.com/Shizoqua) و[@zeus229](https://github.com/zeus229) و[@shadowinlife](https://github.com/shadowinlife)!- **2026-09-23** 🇦🇷 **أسهم أرجنتينية، ومحفظة بيع على المكشوف كان حجمها مقلوبًا، والنصف الأخير من بوابة مخاطر الذيل**: رموز `.BA` (أسهم BYMA وشهادات CEDEAR) تُوجَّه إلى Yahoo كسوق بعملة ARS خاصة بها، وواجهة الويب تجمعها وتسمّيها مثل أي سوق آخر؛ ويظل الاختبار الرجعي للأرجنتين يفشل صراحةً حتى تُنمذَج قواعد التنفيذ في BYMA ([#1543](https://github.com/HKUDS/Vibe-Trading/pull/1543)). كان مُحسِّنا المتوسط-التباين ومراعاة الدوران يقيّمان كل مركز كأنه شراء، فحصلت أقوى مراكز البيع على أقل رأس مال. الاثنان الآن يحدّدان الحجم على عوائد المركز، حيث يتحوّط شراء وبيع لورقتين مترابطتين بدل أن يُقرآ كمترابطتين — على زوج بترابط +0.92 كان الهدف القديم يضع المحفظة كلها في الشراء، والجديد يقسمها 0.51 / -0.49 ([#1548](https://github.com/HKUDS/Vibe-Trading/pull/1548)). وصار رقم مخاطر الذيل مُلزَمًا بتسمية حقله بعد أن تحتفظ الجلسة بأكثر من مقياس أو مستوى ثقة، فلا يمكن لـ ES 95% أن يقتبس قيمة VaR 95% تحت معرّف نداء ([#1425](https://github.com/HKUDS/Vibe-Trading/issues/1425)). **إصلاحات:** تحقّق مونت كارلو يحوّل إلى أساس سنوي بعدد أشرطة المنصة لا بالرقم الثابت 252، فيتوافق شارب مع بقية التقرير نفسه ([#1546](https://github.com/HKUDS/Vibe-Trading/pull/1546)); أمر العقود المستقبلية على شريط بسعر سالب لم يعد يُرفض كليًا ([#1547](https://github.com/HKUDS/Vibe-Trading/pull/1547)); ونسبة IC للعامل تُترك غير محددة عندما يكون الأساس سالبًا بدل قراءة التدهور الحقيقي كتحسّن سليم ([#1549](https://github.com/HKUDS/Vibe-Trading/pull/1549)); وذاكرتان تتشاركان العنوان تحت نوعين مختلفين تحتفظان بسطري فهرس ووصلات مستقلة ([#1545](https://github.com/HKUDS/Vibe-Trading/pull/1545)). شكرًا [@zeus229](https://github.com/zeus229) و[@Shizoqua](https://github.com/Shizoqua) و[@he-yufeng](https://github.com/he-yufeng)!diff --git a/README_es.md b/README_es.mdindex d8815abc3..b6629fbcb 100644--- a/README_es.md+++ b/README_es.md@@ -52,20 +52,22 @@> ⚠️ **Advertencia de seguridad:** la cuenta de X `VibeTrading_HKU`, el proyecto de Virtuals `101845` y el contrato de token `0x640BDBF77b6447E8b7DB7894cED84BD1c40571f4` no son activos oficiales de Vibe-Trading. Nunca hemos lanzado ni respaldado ningún token o memecoin. No compres, conectes una wallet ni firmes nada. [Detalles](SECURITY.md#official-channels--impersonation).-- **2026-09-29** 🛠️ **Las comparaciones largas terminan y los informes de backtest conservan sus cifras**: La compactación de la conversación se ajusta ahora a la ventana de contexto real del modelo (272K para GPT-6 en Codex, 1M para DeepSeek V4) en lugar de una estimación fija de 40K que empezaba a descartar resultados de herramientas desde el tercer paso, de modo que una comparación de 12 empresas volvía a descargar los mismos estados financieros hasta detenerse por falta de progreso. Las ejecuciones fallidas conservan sus pasos y el mensaje de parada indica la herramienta y el motivo; en Codex, las instrucciones añadidas a mitad de la ejecución ya no sustituyen el prompt del sistema, el razonamiento se conserva entre turnos y el modelo predeterminado es `gpt-6-sol`. La verificación de cifras lee ahora la salida del propio backtest (Sortino, Calmar, rotación, pesos, p-valores de Monte Carlo) y acepta una cifra que nombra la ejecución de la que procede, así que un informe de paridad de riesgo frente a pesos iguales ya no se sustituye por un rechazo. `get_market_data` pide `AAPL.US` en lugar de devolver datos vacíos para un ticker estadounidense sin sufijo, y un backtest multimercado recurre a fuentes alternativas por mercado.+- **2026-09-29** 🚀 **Publicada la v0.1.16** ([Notas de la versión](https://github.com/HKUDS/Vibe-Trading/releases/tag/v0.1.16), `pip install -U vibe-trading-ai`): 492 commits y 116 pull requests fusionados desde la 0.1.15, de 16 colaboradores. **El tema de este ciclo es que cada cifra pueda mostrar de dónde viene.** La verificación de cifras ya no adivina qué es un número a partir de las palabras que lo rodean: el modelo declara el papel de cada cifra (observed, derived, proposed, cited o count), la verificación la contrasta con los resultados de las herramientas de la sesión y una cifra que no pasa se elimina en lugar de rechazar la respuesta entera. La salida del propio backtest — Sortino, rotación, pesos y p-valores de Monte Carlo, no solo el Sharpe — sirve ahora de evidencia para el informe sobre él, y las fichas de ejecución citan una métrica solo cuando coincide con el CSV guardado. Las investigaciones largas ya no pierden lo que obtuvieron: la compactación sigue la ventana de contexto real de cada modelo en lugar de una estimación fija de 40K. La revisión de datos faltantes alcanzó 59 alphas más. **Corregido a partir del informe de un usuario:** las ejecuciones fallidas conservan sus pasos y el mensaje de parada indica la llamada que no produjo nada; `read_file` dice dónde puede leer en lugar de dejar que el modelo adivine rutas hasta detenerse por falta de progreso; en Codex, las instrucciones añadidas a mitad de la ejecución ya no sustituyen el prompt del sistema, el razonamiento se conserva entre turnos y el modelo predeterminado es `gpt-6-sol`; y `get_market_data` pide `AAPL.US` en lugar de devolver datos vacíos para un ticker estadounidense sin sufijo. **Novedades:** Argentina (BYMA), velas semanales y mensuales, bahasa indonesio, Gildata, OpenCode, canales de correo electrónico y WebSocket, configuración de canales de mensajería desde la interfaz web, y conectores para KIS, Upbit, Toss Securities y Scalable Capital, 18 brókeres en total. ¡Gracias a [@Shizoqua](https://github.com/Shizoqua), [@zeus229](https://github.com/zeus229), [@cgycorey](https://github.com/cgycorey), [@shadowinlife](https://github.com/shadowinlife), [@lorenzozanee](https://github.com/lorenzozanee), [@he-yufeng](https://github.com/he-yufeng), [@chiww](https://github.com/chiww), [@as950118](https://github.com/as950118), [@woshi77777stars](https://github.com/woshi77777stars), [@Yoruxyv](https://github.com/Yoruxyv), [@sambazhu](https://github.com/sambazhu), [@0xouzm](https://github.com/0xouzm), [@tingkk](https://github.com/tingkk), [@tonydo](https://github.com/tonydo), [@alanwilhelm](https://github.com/alanwilhelm) y [@modelpath-dev](https://github.com/modelpath-dev)!- **2026-09-28** 🛠️ **Estado de fuentes y capacidades de brókeres**: las sondas periódicas de fuentes públicas, sin credenciales y con tiempo limitado, notifican fallos y anomalías de datos ([#1627](https://github.com/HKUDS/Vibe-Trading/pull/1627)). BaoStock incorpora plazos de comunicación y serializa sesiones concurrentes ([#1615](https://github.com/HKUDS/Vibe-Trading/pull/1615)); las consultas de credenciales de Copilot se renuevan tras una caché breve ([#1619](https://github.com/HKUDS/Vibe-Trading/pull/1619)). La matriz generada separa permisos de simulación y operaciones reales por perfil, sin confundir declaraciones con verificación práctica ([#1629](https://github.com/HKUDS/Vibe-Trading/pull/1629)). **Mejoras de verificación y MT5**: se conservan los símbolos explícitos y las referencias exactas a campos; los números con separadores locales mantienen la tolerancia existente ([#1584](https://github.com/HKUDS/Vibe-Trading/pull/1584), [#1586](https://github.com/HKUDS/Vibe-Trading/pull/1586), [#1588](https://github.com/HKUDS/Vibe-Trading/pull/1588)). Las correcciones de solo texto tienen un límite de intentos ([#1600](https://github.com/HKUDS/Vibe-Trading/pull/1600)). MT5 busca en el terminal seleccionado y rechaza alias ambiguos; el entorno aislado del backtest solo recibe ajustes de conexión validados ([#1597](https://github.com/HKUDS/Vibe-Trading/pull/1597), [#1598](https://github.com/HKUDS/Vibe-Trading/pull/1598)).- **2026-09-27** 🛠️ **Tarjetas verificables y mejoras en el análisis**: JSON, Markdown y Run Detail muestran registros hash de ejecución del backtest y referencias verificadas a métricas CSV ([#1612](https://github.com/HKUDS/Vibe-Trading/pull/1612)). Se corrigen el recuento de días y los empates en factores GTJA ([#1604](https://github.com/HKUDS/Vibe-Trading/pull/1604)), y el cierre de cortos en India comprueba el límite de precio del lado comprador ([#1608](https://github.com/HKUDS/Vibe-Trading/pull/1608)). El uso de tokens sigue visible sin permitir cadenas de credenciales arbitrarias ([#1606](https://github.com/HKUDS/Vibe-Trading/pull/1606)); se recuperan las propiedades de esquemas MCP malformados ([#1607](https://github.com/HKUDS/Vibe-Trading/pull/1607)). El filtro de órdenes inmediatas permite preguntas de análisis ([#1605](https://github.com/HKUDS/Vibe-Trading/pull/1605)), y la documentación identifica OpenRouter como configuración predeterminada ([#1609](https://github.com/HKUDS/Vibe-Trading/issues/1609)).+<details>+<summary>Noticias anteriores</summary>+- **2026-09-26** 🛠️ **Indicadores y límites de negociación más fiables**: Los indicadores técnicos conservan la fecha de observación y pueden reutilizar resultados eliminados al compactar la conversación ([#1590](https://github.com/HKUDS/Vibe-Trading/pull/1590), [#1601](https://github.com/HKUDS/Vibe-Trading/pull/1601)). Las cinco ventanas WVMA de Qlib158 usan rendimientos absolutos en el numerador ([#1594](https://github.com/HKUDS/Vibe-Trading/pull/1594)); la comprobación de Benford conserva el primer dígito correcto en los límites numéricos ([#1591](https://github.com/HKUDS/Vibe-Trading/pull/1591)). Un límite de exposición fijado explícitamente en cero permanece en cero ([#1593](https://github.com/HKUDS/Vibe-Trading/pull/1593)), y las órdenes simuladas de Dhan rechazan cantidades fraccionarias o inválidas sin perder precisión por redondeo ([#1595](https://github.com/HKUDS/Vibe-Trading/pull/1595)). Las variables de entorno numéricas opcionales inválidas vuelven a sus valores predeterminados, manteniendo las restricciones de la configuración explícita ([#1592](https://github.com/HKUDS/Vibe-Trading/pull/1592)). Los activadores de ejecución conservan los valores predeterminados de sus campos y el comportamiento correcto de sus métodos de creación en subclases ([#1599](https://github.com/HKUDS/Vibe-Trading/pull/1599)).- **2026-09-25** 🛠️ **Sesiones de análisis y diagnóstico de canales más fiables**: Una columna de texto en el CSV de un backtest ya no descarta las métricas numéricas ([#1579](https://github.com/HKUDS/Vibe-Trading/pull/1579)), y la compactación de conversaciones incluye el contenido de razonamiento en el presupuesto de mensajes conservados ([#1582](https://github.com/HKUDS/Vibe-Trading/pull/1582)). Los fallos al cargar canales registran la excepción real ([#1580](https://github.com/HKUDS/Vibe-Trading/pull/1580)); la guía de Telegram aclara que los controles CLI y Web comparten el mismo entorno de ejecución de la API ([#1583](https://github.com/HKUDS/Vibe-Trading/pull/1583)). La lectura de carteras de Robinhood rechaza objetos de poder de compra con formato incorrecto y mantiene como desconocidos los valores omitidos o null ([#1526](https://github.com/HKUDS/Vibe-Trading/pull/1526)). Al salir del chat también se cancelan los temporizadores pendientes de desplazamiento del historial. ¡Gracias a [@Shizoqua](https://github.com/Shizoqua) y [@lorenzozanee](https://github.com/lorenzozanee)!-- **2026-09-24** 💱 **Una línea en dólares valuada en pesos, Email y WebSocket en la Web UI, y una rotura de CI por una actualización del SDK**: BYMA y la TSX listan líneas en dólares junto a las de su moneda local (GGALD.BA, DLR-U.TO), y un backtest argentino o canadiense las valuaba en su único pool en pesos o en dólares canadienses. Los loaders ahora admiten una línea solo en la moneda de su mercado, como ya ocurría con las líneas de la LSE, y cualquier otra moneda de cotización declarada viaja con los datos, de modo que la respuesta nombra la correcta ([#1576](https://github.com/HKUDS/Vibe-Trading/pull/1576)). Las líneas de Hong Kong en renminbi y en dólares (80700.HK, 9834.HK), que HKEX numera en sus propios rangos de códigos, se rechazan en un backtest de Hong Kong por la misma razón y se cotizan en su propia moneda. Los perfiles de acciones incluyen el mercado y la moneda del propio listado junto a los fundamentales del emisor ([#1577](https://github.com/HKUDS/Vibe-Trading/pull/1577)), el grounding y la búsqueda de símbolos conocen todos los mercados que enruta la capa de datos ([#1575](https://github.com/HKUDS/Vibe-Trading/pull/1575)) y `technical_indicators` informa el volumen con la unidad que declara su fuente — lotes o acciones, a 100× de distancia ([#1571](https://github.com/HKUDS/Vibe-Trading/pull/1571)). Email y WebSocket se suman a la configuración guiada de canales, con verificación de certificados del servidor de correo y escrituras de configuración cerradas a páginas de otros sitios, y un inicio de sesión IMAP sin SSL que ahora se cifra con STARTTLS antes de enviar la contraseña ([#1544](https://github.com/HKUDS/Vibe-Trading/pull/1544)). **Corregido:** una cabecera explícita del proveedor que perdía frente a su gemela del entorno con openai 3.19.2 y había puesto en rojo el CI de todos los PR ([#1568](https://github.com/HKUDS/Vibe-Trading/pull/1568)); un backtest de una fuente con resolución de microsegundos fechado en 1970 ([#1560](https://github.com/HKUDS/Vibe-Trading/pull/1560)); un objetivo de investigación que es claramente una orden vuelve a rechazarse ([#1562](https://github.com/HKUDS/Vibe-Trading/pull/1562)); una división OOS del strict bench que podía dejar un lado vacío ([#1559](https://github.com/HKUDS/Vibe-Trading/pull/1559)); alineación de etiquetas en HRP, orden de etiquetas en la validación cruzada y entradas no finitas ([#1555](https://github.com/HKUDS/Vibe-Trading/pull/1555), [#1556](https://github.com/HKUDS/Vibe-Trading/pull/1556), [#1557](https://github.com/HKUDS/Vibe-Trading/pull/1557), [#1558](https://github.com/HKUDS/Vibe-Trading/pull/1558)); la ventana de sobreoperación de la cuenta sombra ([#1563](https://github.com/HKUDS/Vibe-Trading/pull/1563)); y el crecimiento de activos, que se medía día contra día ([#1564](https://github.com/HKUDS/Vibe-Trading/pull/1564)). ¡Gracias [@Shizoqua](https://github.com/Shizoqua), [@zeus229](https://github.com/zeus229) y [@shadowinlife](https://github.com/shadowinlife)!-<details>-<summary>Noticias anteriores</summary>++- **2026-09-24** 💱 **Una línea en dólares valuada en pesos, Email y WebSocket en la Web UI, y una rotura de CI por una actualización del SDK**: BYMA y la TSX listan líneas en dólares junto a las de su moneda local (GGALD.BA, DLR-U.TO), y un backtest argentino o canadiense las valuaba en su único pool en pesos o en dólares canadienses. Los loaders ahora admiten una línea solo en la moneda de su mercado, como ya ocurría con las líneas de la LSE, y cualquier otra moneda de cotización declarada viaja con los datos, de modo que la respuesta nombra la correcta ([#1576](https://github.com/HKUDS/Vibe-Trading/pull/1576)). Las líneas de Hong Kong en renminbi y en dólares (80700.HK, 9834.HK), que HKEX numera en sus propios rangos de códigos, se rechazan en un backtest de Hong Kong por la misma razón y se cotizan en su propia moneda. Los perfiles de acciones incluyen el mercado y la moneda del propio listado junto a los fundamentales del emisor ([#1577](https://github.com/HKUDS/Vibe-Trading/pull/1577)), el grounding y la búsqueda de símbolos conocen todos los mercados que enruta la capa de datos ([#1575](https://github.com/HKUDS/Vibe-Trading/pull/1575)) y `technical_indicators` informa el volumen con la unidad que declara su fuente — lotes o acciones, a 100× de distancia ([#1571](https://github.com/HKUDS/Vibe-Trading/pull/1571)). Email y WebSocket se suman a la configuración guiada de canales, con verificación de certificados del servidor de correo y escrituras de configuración cerradas a páginas de otros sitios, y un inicio de sesión IMAP sin SSL que ahora se cifra con STARTTLS antes de enviar la contraseña ([#1544](https://github.com/HKUDS/Vibe-Trading/pull/1544)). **Corregido:** una cabecera explícita del proveedor que perdía frente a su gemela del entorno con openai 3.19.2 y había puesto en rojo el CI de todos los PR ([#1568](https://github.com/HKUDS/Vibe-Trading/pull/1568)); un backtest de una fuente con resolución de microsegundos fechado en 1970 ([#1560](https://github.com/HKUDS/Vibe-Trading/pull/1560)); un objetivo de investigación que es claramente una orden vuelve a rechazarse ([#1562](https://github.com/HKUDS/Vibe-Trading/pull/1562)); una división OOS del strict bench que podía dejar un lado vacío ([#1559](https://github.com/HKUDS/Vibe-Trading/pull/1559)); alineación de etiquetas en HRP, orden de etiquetas en la validación cruzada y entradas no finitas ([#1555](https://github.com/HKUDS/Vibe-Trading/pull/1555), [#1556](https://github.com/HKUDS/Vibe-Trading/pull/1556), [#1557](https://github.com/HKUDS/Vibe-Trading/pull/1557), [#1558](https://github.com/HKUDS/Vibe-Trading/pull/1558)); la ventana de sobreoperación de la cuenta sombra ([#1563](https://github.com/HKUDS/Vibe-Trading/pull/1563)); y el crecimiento de activos, que se medía día contra día ([#1564](https://github.com/HKUDS/Vibe-Trading/pull/1564)). ¡Gracias [@Shizoqua](https://github.com/Shizoqua), [@zeus229](https://github.com/zeus229) y [@shadowinlife](https://github.com/shadowinlife)!- **2026-09-23** 🇦🇷 **Cotizaciones argentinas, un libro de cortos dimensionado al revés y la última mitad de la puerta de riesgo de cola**: los símbolos `.BA` —acciones de BYMA y CEDEARs— se enrutan a Yahoo como mercado propio en ARS, y la interfaz web los agrupa y etiqueta como a cualquier otro; un backtest argentino sigue fallando de forma explícita hasta que se modelen las reglas de ejecución de BYMA ([#1543](https://github.com/HKUDS/Vibe-Trading/pull/1543)). Los optimizadores de media-varianza y de rotación puntuaban cada posición como si fuera larga, así que los mejores cortos recibían el menor capital. Ahora ambos dimensionan sobre retornos de la posición, donde un largo y un corto de dos nombres correlacionados se cubren entre sí en vez de leerse como correlacionados: en un par con correlación +0,92 el objetivo anterior ponía todo el libro en el largo y el nuevo lo reparte 0,51 / -0,49 ([#1548](https://github.com/HKUDS/Vibe-Trading/pull/1548)). Y una cifra de riesgo de cola debe nombrar su propio campo en cuanto una sesión contiene más de una medida o nivel de confianza, de modo que un ES 95% ya no puede citar el valor del VaR 95% bajo el id de una llamada ([#1425](https://github.com/HKUDS/Vibe-Trading/issues/1425)). **Corregido:** la validación Monte Carlo anualiza con las barras por año del mercado en lugar de un 252 fijo, así su Sharpe concuerda con el resto del mismo informe ([#1546](https://github.com/HKUDS/Vibe-Trading/pull/1546)); una orden de futuros en una barra de precio negativo ya no se rechaza sin más ([#1547](https://github.com/HKUDS/Vibe-Trading/pull/1547)); el IC ratio de un factor queda sin definir cuando la línea base es negativa, en vez de leer un deterioro real como mejora ([#1549](https://github.com/HKUDS/Vibe-Trading/pull/1549)); y dos memorias que comparten título bajo tipos distintos conservan sus propias filas de índice y enlaces ([#1545](https://github.com/HKUDS/Vibe-Trading/pull/1545)). ¡Gracias [@zeus229](https://github.com/zeus229), [@Shizoqua](https://github.com/Shizoqua) y [@he-yufeng](https://github.com/he-yufeng)!diff --git a/README_id.md b/README_id.mdindex 49b4a5768..1609a4149 100644--- a/README_id.md+++ b/README_id.md@@ -53,20 +53,22 @@> ⚠️ **Peringatan keamanan:** Akun X `VibeTrading_HKU`, proyek Virtuals `101845`, dan kontrak token `0x640BDBF77b6447E8b7DB7894cED84BD1c40571f4` bukan aset resmi Vibe-Trading. Kami tidak pernah meluncurkan atau mendukung token maupun memecoin apa pun. Jangan membeli, menghubungkan wallet, atau menandatangani apa pun. [Detail](SECURITY.md#official-channels--impersonation).-- **2026-09-29** 🛠️ **Perbandingan panjang selesai, dan laporan backtest mempertahankan angkanya**: Pemadatan percakapan kini disesuaikan dengan jendela konteks model yang sebenarnya (272K untuk GPT-6 di Codex, 1M untuk DeepSeek V4), bukan estimasi tetap 40K yang mulai membuang hasil alat sejak langkah ketiga sehingga perbandingan 12 perusahaan mengambil ulang laporan keuangan yang sama sampai berhenti karena tidak ada kemajuan. Eksekusi yang gagal tetap menyimpan langkahnya dan pesan penghentian menyebutkan alat serta alasannya; di Codex, instruksi yang ditambahkan di tengah eksekusi tidak lagi menggantikan prompt sistem, penalaran dibawa antar giliran, dan model bawaan adalah `gpt-6-sol`. Pemeriksaan angka kini membaca keluaran backtest itu sendiri (Sortino, Calmar, perputaran, bobot, nilai p Monte Carlo) dan menerima angka yang menyebut eksekusi asalnya, sehingga laporan risk parity versus bobot setara tidak lagi diganti dengan penolakan. `get_market_data` meminta `AAPL.US` alih-alih mengembalikan data kosong untuk ticker AS tanpa sufiks, dan backtest multi-pasar beralih ke sumber cadangan per pasar.+- **2026-09-29** 🚀 **v0.1.16 dirilis** ([Catatan rilis](https://github.com/HKUDS/Vibe-Trading/releases/tag/v0.1.16), `pip install -U vibe-trading-ai`): 492 commit dan 116 pull request yang digabungkan sejak 0.1.15, dari 16 kontributor. **Tema siklus ini adalah angka yang dapat menunjukkan dari mana asalnya.** Pemeriksaan angka tidak lagi menebak arti sebuah angka dari kata-kata di sekitarnya: model menyatakan peran setiap angka (observed, derived, proposed, cited, atau count), pemeriksaan mencocokkannya dengan hasil alat dalam sesi tersebut, dan angka yang gagal dihapus alih-alih seluruh jawaban ditolak. Keluaran backtest itu sendiri — Sortino, perputaran, bobot, dan nilai p Monte Carlo, bukan hanya Sharpe — kini menjadi bukti bagi laporan tentangnya, dan kartu eksekusi hanya mengutip metrik yang cocok dengan CSV yang tersimpan. Riset panjang tidak lagi kehilangan data yang telah diambil: pemadatan konteks mengikuti jendela konteks sebenarnya dari setiap model, bukan estimasi tetap 40K. Pemeriksaan data yang hilang menjangkau 59 alpha lagi. **Diperbaiki dari laporan pengguna:** eksekusi yang gagal tetap menyimpan langkahnya dan pesan penghentian menyebutkan panggilan yang tidak menghasilkan apa pun; `read_file` menjelaskan di mana ia dapat membaca, sehingga model tidak lagi menebak jalur sampai berhenti karena tidak ada kemajuan; di Codex, instruksi di tengah eksekusi tidak lagi menggantikan prompt sistem, penalaran dibawa antar giliran, dan model bawaan adalah `gpt-6-sol`; dan `get_market_data` meminta `AAPL.US` alih-alih mengembalikan data kosong untuk ticker AS tanpa sufiks. **Baru:** Argentina (BYMA), bar mingguan dan bulanan, Bahasa Indonesia, Gildata, OpenCode, kanal email dan WebSocket, pengaturan kanal pesan dari UI web, serta konektor KIS, Upbit, Toss Securities, dan Scalable Capital, total 18 broker. Terima kasih [@Shizoqua](https://github.com/Shizoqua), [@zeus229](https://github.com/zeus229), [@cgycorey](https://github.com/cgycorey), [@shadowinlife](https://github.com/shadowinlife), [@lorenzozanee](https://github.com/lorenzozanee), [@he-yufeng](https://github.com/he-yufeng), [@chiww](https://github.com/chiww), [@as950118](https://github.com/as950118), [@woshi77777stars](https://github.com/woshi77777stars), [@Yoruxyv](https://github.com/Yoruxyv), [@sambazhu](https://github.com/sambazhu), [@0xouzm](https://github.com/0xouzm), [@tingkk](https://github.com/tingkk), [@tonydo](https://github.com/tonydo), [@alanwilhelm](https://github.com/alanwilhelm), dan [@modelpath-dev](https://github.com/modelpath-dev)!- **2026-09-28** 🛠️ **Kesehatan sumber data dan transparansi broker**: pemeriksaan berkala sumber publik tanpa kredensial dan dengan batas waktu melaporkan gangguan koneksi serta anomali data ([#1627](https://github.com/HKUDS/Vibe-Trading/pull/1627)). BaoStock menambahkan tenggat komunikasi dan mengurutkan sesi bersamaan ([#1615](https://github.com/HKUDS/Vibe-Trading/pull/1615)); pencarian kredensial Copilot diperbarui setelah cache singkat berakhir ([#1619](https://github.com/HKUDS/Vibe-Trading/pull/1619)). Matriks broker yang dihasilkan otomatis memisahkan izin simulasi dan perdagangan riil per profil, serta membedakan kemampuan yang dideklarasikan dari verifikasi runtime ([#1629](https://github.com/HKUDS/Vibe-Trading/pull/1629)). **Perbaikan verifikasi dan MT5**: simbol eksplisit dan referensi bidang yang tepat dipertahankan; pemisah angka lokal tetap menggunakan toleransi numerik yang ada ([#1584](https://github.com/HKUDS/Vibe-Trading/pull/1584), [#1586](https://github.com/HKUDS/Vibe-Trading/pull/1586), [#1588](https://github.com/HKUDS/Vibe-Trading/pull/1588)). Percobaan koreksi khusus teks dibatasi ([#1600](https://github.com/HKUDS/Vibe-Trading/pull/1600)). MT5 mencari melalui terminal terpilih dan menolak alias broker yang ambigu; sandbox backtest hanya menerima pengaturan koneksi yang tervalidasi ([#1597](https://github.com/HKUDS/Vibe-Trading/pull/1597), [#1598](https://github.com/HKUDS/Vibe-Trading/pull/1598)).- **2026-09-27** 🛠️ **Kartu eksekusi yang dapat diverifikasi dan perbaikan riset**: JSON, Markdown, dan Run Detail menampilkan catatan hash eksekusi backtest serta referensi metrik CSV yang terverifikasi ([#1612](https://github.com/HKUDS/Vibe-Trading/pull/1612)). Hitungan hari dan nilai ekstrem seri pada faktor GTJA diperbaiki ([#1604](https://github.com/HKUDS/Vibe-Trading/pull/1604)); penutupan posisi short India memeriksa batas harga sisi beli ([#1608](https://github.com/HKUDS/Vibe-Trading/pull/1608)). Penggunaan token tetap terbaca tanpa meloloskan string kredensial sembarang ([#1606](https://github.com/HKUDS/Vibe-Trading/pull/1606)), properti skema MCP yang rusak dipulihkan ([#1607](https://github.com/HKUDS/Vibe-Trading/pull/1607)), dan pemeriksaan tujuan pesanan langsung tetap mengizinkan pertanyaan riset ([#1605](https://github.com/HKUDS/Vibe-Trading/pull/1605)). Dokumentasi kini menyatakan OpenRouter sebagai pengaturan bawaan ([#1609](https://github.com/HKUDS/Vibe-Trading/issues/1609)).+<details>+<summary>Berita sebelumnya</summary>+- **2026-09-26** 🛠️ **Indikator dan batas perdagangan lebih andal**: Indikator teknis mempertahankan tanggal pengamatan dan dapat memakai kembali hasil yang terhapus saat percakapan dipadatkan ([#1590](https://github.com/HKUDS/Vibe-Trading/pull/1590), [#1601](https://github.com/HKUDS/Vibe-Trading/pull/1601)). Kelima jendela WVMA Qlib158 memakai imbal hasil absolut pada pembilang ([#1594](https://github.com/HKUDS/Vibe-Trading/pull/1594)); pemeriksaan Benford mempertahankan digit pertama yang benar pada batas numerik ([#1591](https://github.com/HKUDS/Vibe-Trading/pull/1591)). Batas eksposur yang secara eksplisit ditetapkan nol tetap nol ([#1593](https://github.com/HKUDS/Vibe-Trading/pull/1593)), dan pesanan simulasi Dhan menolak jumlah pecahan atau tidak valid tanpa kehilangan presisi akibat pembulatan ([#1595](https://github.com/HKUDS/Vibe-Trading/pull/1595)). Variabel lingkungan numerik opsional yang tidak valid kembali ke nilai bawaan, sementara batasan konfigurasi eksplisit tetap berlaku ([#1592](https://github.com/HKUDS/Vibe-Trading/pull/1592)). Pemicu runtime juga mempertahankan nilai bawaan kolom dan perilaku metode pembuat objek pada subkelas dengan benar ([#1599](https://github.com/HKUDS/Vibe-Trading/pull/1599)).- **2026-09-25** 🛠️ **Sesi riset dan diagnostik channel lebih andal**: Kolom teks dalam CSV backtest tidak lagi menghilangkan seluruh metrik numerik ([#1579](https://github.com/HKUDS/Vibe-Trading/pull/1579)), dan pemadatan percakapan kini menghitung konten penalaran dalam anggaran pesan yang dipertahankan ([#1582](https://github.com/HKUDS/Vibe-Trading/pull/1582)). Kegagalan memuat channel mencatat pengecualian yang sebenarnya ([#1580](https://github.com/HKUDS/Vibe-Trading/pull/1580)); panduan Telegram memperjelas bahwa kontrol CLI dan Web berbagi runtime API yang sama ([#1583](https://github.com/HKUDS/Vibe-Trading/pull/1583)). Pembacaan portofolio Robinhood menolak objek daya beli dengan format yang salah, sementara nilai yang tidak disertakan atau null tetap dinyatakan tidak diketahui ([#1526](https://github.com/HKUDS/Vibe-Trading/pull/1526)). Meninggalkan halaman chat juga membatalkan timer pengguliran riwayat yang masih tertunda. Terima kasih [@Shizoqua](https://github.com/Shizoqua) dan [@lorenzozanee](https://github.com/lorenzozanee)!-- **2026-09-24** 💱 **Saham berdenominasi dolar yang dihitung dalam peso, Email dan WebSocket di Web UI, dan CI yang rusak karena pembaruan SDK**: BYMA dan TSX mencatatkan saham berdenominasi dolar AS di samping saham bermata uang lokal (GGALD.BA, DLR-U.TO), dan backtest Argentina atau Kanada menilainya dalam satu-satunya pool peso atau dolar Kanada. Loader kini hanya menerima saham yang dikutip dalam mata uang pasarnya, seperti yang sudah berlaku untuk saham LSE, dan mata uang kuotasi lain yang dideklarasikan sumber ikut terbawa bersama data sehingga jawaban menyebut mata uang yang benar ([#1576](https://github.com/HKUDS/Vibe-Trading/pull/1576)). Counter RMB dan USD Hong Kong (80700.HK, 9834.HK), yang diberi nomor HKEX dalam rentang kodenya sendiri, ditolak dalam backtest Hong Kong karena alasan yang sama dan dikutip dalam mata uangnya sendiri. Profil saham membawa bursa dan mata uang pencatatan itu sendiri di samping fundamental penerbit ([#1577](https://github.com/HKUDS/Vibe-Trading/pull/1577)), grounding dan pencarian simbol mengenali setiap pasar yang dirutekan lapisan data ([#1575](https://github.com/HKUDS/Vibe-Trading/pull/1575)), dan `technical_indicators` melaporkan volume beserta satuan yang dideklarasikan sumbernya — lot atau lembar saham, selisih 100× ([#1571](https://github.com/HKUDS/Vibe-Trading/pull/1571)). Email dan WebSocket bergabung ke pengaturan channel terpandu, dengan sertifikat server email kini diverifikasi dan penulisan pengaturan tertutup bagi halaman lintas situs, serta login IMAP tanpa SSL yang kini dienkripsi dengan STARTTLS sebelum kata sandi dikirim ([#1544](https://github.com/HKUDS/Vibe-Trading/pull/1544)). **Diperbaiki:** header provider eksplisit yang kalah oleh header lingkungan bernama sama di openai 3.19.2 dan sempat membuat CI setiap PR merah ([#1568](https://github.com/HKUDS/Vibe-Trading/pull/1568)); backtest dari sumber beresolusi mikrodetik yang bertanggal 1970 ([#1560](https://github.com/HKUDS/Vibe-Trading/pull/1560)); target riset yang jelas-jelas berupa order kembali ditolak ([#1562](https://github.com/HKUDS/Vibe-Trading/pull/1562)); pembagian OOS strict bench yang bisa membuat satu sisi kosong ([#1559](https://github.com/HKUDS/Vibe-Trading/pull/1559)); penyelarasan label HRP, urutan label validasi silang, dan input kuant yang tidak hingga ([#1555](https://github.com/HKUDS/Vibe-Trading/pull/1555), [#1556](https://github.com/HKUDS/Vibe-Trading/pull/1556), [#1557](https://github.com/HKUDS/Vibe-Trading/pull/1557), [#1558](https://github.com/HKUDS/Vibe-Trading/pull/1558)); jendela overtrading akun bayangan ([#1563](https://github.com/HKUDS/Vibe-Trading/pull/1563)); serta pertumbuhan aset yang dihitung hari ke hari ([#1564](https://github.com/HKUDS/Vibe-Trading/pull/1564)). Terima kasih [@Shizoqua](https://github.com/Shizoqua), [@zeus229](https://github.com/zeus229), dan [@shadowinlife](https://github.com/shadowinlife)!-<details>-<summary>Berita sebelumnya</summary>++- **2026-09-24** 💱 **Saham berdenominasi dolar yang dihitung dalam peso, Email dan WebSocket di Web UI, dan CI yang rusak karena pembaruan SDK**: BYMA dan TSX mencatatkan saham berdenominasi dolar AS di samping saham bermata uang lokal (GGALD.BA, DLR-U.TO), dan backtest Argentina atau Kanada menilainya dalam satu-satunya pool peso atau dolar Kanada. Loader kini hanya menerima saham yang dikutip dalam mata uang pasarnya, seperti yang sudah berlaku untuk saham LSE, dan mata uang kuotasi lain yang dideklarasikan sumber ikut terbawa bersama data sehingga jawaban menyebut mata uang yang benar ([#1576](https://github.com/HKUDS/Vibe-Trading/pull/1576)). Counter RMB dan USD Hong Kong (80700.HK, 9834.HK), yang diberi nomor HKEX dalam rentang kodenya sendiri, ditolak dalam backtest Hong Kong karena alasan yang sama dan dikutip dalam mata uangnya sendiri. Profil saham membawa bursa dan mata uang pencatatan itu sendiri di samping fundamental penerbit ([#1577](https://github.com/HKUDS/Vibe-Trading/pull/1577)), grounding dan pencarian simbol mengenali setiap pasar yang dirutekan lapisan data ([#1575](https://github.com/HKUDS/Vibe-Trading/pull/1575)), dan `technical_indicators` melaporkan volume beserta satuan yang dideklarasikan sumbernya — lot atau lembar saham, selisih 100× ([#1571](https://github.com/HKUDS/Vibe-Trading/pull/1571)). Email dan WebSocket bergabung ke pengaturan channel terpandu, dengan sertifikat server email kini diverifikasi dan penulisan pengaturan tertutup bagi halaman lintas situs, serta login IMAP tanpa SSL yang kini dienkripsi dengan STARTTLS sebelum kata sandi dikirim ([#1544](https://github.com/HKUDS/Vibe-Trading/pull/1544)). **Diperbaiki:** header provider eksplisit yang kalah oleh header lingkungan bernama sama di openai 3.19.2 dan sempat membuat CI setiap PR merah ([#1568](https://github.com/HKUDS/Vibe-Trading/pull/1568)); backtest dari sumber beresolusi mikrodetik yang bertanggal 1970 ([#1560](https://github.com/HKUDS/Vibe-Trading/pull/1560)); target riset yang jelas-jelas berupa order kembali ditolak ([#1562](https://github.com/HKUDS/Vibe-Trading/pull/1562)); pembagian OOS strict bench yang bisa membuat satu sisi kosong ([#1559](https://github.com/HKUDS/Vibe-Trading/pull/1559)); penyelarasan label HRP, urutan label validasi silang, dan input kuant yang tidak hingga ([#1555](https://github.com/HKUDS/Vibe-Trading/pull/1555), [#1556](https://github.com/HKUDS/Vibe-Trading/pull/1556), [#1557](https://github.com/HKUDS/Vibe-Trading/pull/1557), [#1558](https://github.com/HKUDS/Vibe-Trading/pull/1558)); jendela overtrading akun bayangan ([#1563](https://github.com/HKUDS/Vibe-Trading/pull/1563)); serta pertumbuhan aset yang dihitung hari ke hari ([#1564](https://github.com/HKUDS/Vibe-Trading/pull/1564)). Terima kasih [@Shizoqua](https://github.com/Shizoqua), [@zeus229](https://github.com/zeus229), dan [@shadowinlife](https://github.com/shadowinlife)!- **2026-09-23** 🇦🇷 **Saham Argentina, buku short yang disizing terbalik, dan separuh terakhir gerbang risiko ekor**: simbol `.BA` — saham BYMA dan CEDEAR — dirutekan ke Yahoo sebagai pasar ARS sendiri, dan Web UI mengelompokkan serta melabelinya seperti pasar lain; backtest Argentina tetap gagal secara eksplisit sampai aturan eksekusi BYMA dimodelkan ([#1543](https://github.com/HKUDS/Vibe-Trading/pull/1543)). Optimizer mean-variance dan turnover-aware menilai setiap posisi seolah long, sehingga short terbaik mendapat modal paling sedikit. Keduanya kini disizing pada return posisi, di mana long dan short dua nama berkorelasi saling melindungi alih-alih terbaca berkorelasi — pada pasangan berkorelasi +0,92 objektif lama menaruh seluruh buku di long, yang baru membaginya 0,51 / -0,49 ([#1548](https://github.com/HKUDS/Vibe-Trading/pull/1548)). Selain itu, angka risiko ekor kini harus menyebut fieldnya sendiri begitu satu sesi memuat lebih dari satu ukuran atau tingkat keyakinan, sehingga ES 95% tidak bisa lagi mengutip nilai VaR 95% dengan id panggilan ([#1425](https://github.com/HKUDS/Vibe-Trading/issues/1425)). **Diperbaiki:** validasi Monte Carlo menyetahunkan dengan bar per tahun bursa, bukan 252 yang dipaku, sehingga Sharpe-nya sejalan dengan bagian lain laporan yang sama ([#1546](https://github.com/HKUDS/Vibe-Trading/pull/1546)); order futures pada bar berharga negatif tidak lagi ditolak seluruhnya ([#1547](https://github.com/HKUDS/Vibe-Trading/pull/1547)); IC ratio sebuah faktor dibiarkan kosong bila baseline-nya negatif, alih-alih membaca pelemahan nyata sebagai perbaikan ([#1549](https://github.com/HKUDS/Vibe-Trading/pull/1549)); dan dua memori dengan judul sama di tipe berbeda mempertahankan baris indeks serta target tautannya masing-masing ([#1545](https://github.com/HKUDS/Vibe-Trading/pull/1545)). Terima kasih [@zeus229](https://github.com/zeus229), [@Shizoqua](https://github.com/Shizoqua), dan [@he-yufeng](https://github.com/he-yufeng)!diff --git a/README_ja.md b/README_ja.mdindex ad4940e62..71d6c5206 100644--- a/README_ja.md+++ b/README_ja.md@@ -52,20 +52,22 @@> ⚠️ **セキュリティ警告:** Xアカウント `VibeTrading_HKU`、Virtualsプロジェクト `101845`、およびトークンコントラクト `0x640BDBF77b6447E8b7DB7894cED84BD1c40571f4` は、いずれもVibe-Trading公式のものではありません。Vibe-Tradingはこれまで、いかなるトークンやミームコインも発行・公認していません。購入、ウォレットの接続、署名は行わないでください。[詳細](SECURITY.md#official-channels--impersonation)-- **2026-09-29** 🛠️ **長い比較が最後まで完了し、バックテストレポートの数値が残るように**:会話の圧縮はモデルの実際のコンテキストウィンドウ(Codex の GPT-6 は 272K、DeepSeek V4 は 1M)に合わせて行うようになりました。従来の固定 40K の推定では 3 ステップ目からツール結果が消され、12 社の比較は同じ財務データを取り直し続けて no-progress で停止していました。失敗した実行もステップを保持し、停止メッセージにはツール名と理由が表示されます。Codex では途中の指示がシステムプロンプトを置き換えなくなり、推論がターン間で引き継がれ、既定モデルは `gpt-6-sol` になりました。数値チェックはバックテスト自身の出力(Sortino、Calmar、回転率、ウェイト、モンテカルロ p 値)を読み、どの実行の値かを示した数値を照合するため、リスクパリティと等ウェイトの比較レポートが拒否文に置き換えられなくなりました。`get_market_data` は接尾辞のない米国株ティッカーに空データを返さず `AAPL.US` を求め、複数市場のバックテストは市場ごとにデータソースをフォールバックします。+- **2026-09-29** 🚀 **v0.1.16 リリース** ([リリースノート](https://github.com/HKUDS/Vibe-Trading/releases/tag/v0.1.16), `pip install -U vibe-trading-ai`):0.1.15 以降のコミット 492 件、マージされたプルリクエスト 116 件、コントリビューター 16 名。**このサイクルのテーマは「どの数値も自分の出どころを示せること」です。** グラウンディングゲートは、数値の周りの言葉からその意味を推測しなくなりました。モデルが各数値の役割(observed / derived / proposed / cited / count)を宣言し、ゲートがそのセッションのツール結果と照合します。照合できない数値は回答全体を拒否するのではなく、その数値だけを削除します。バックテスト自身の出力(シャープレシオだけでなく、ソルティノ、回転率、ウェイト、モンテカルロ p 値)がそのレポートの根拠になり、ランカードは保存された CSV と一致する指標だけを引用します。長い調査タスクは取得したデータを失わなくなりました。コンテキスト圧縮は固定の 40K 推定ではなく、各モデルの実際のコンテキストウィンドウに従います。欠損データの点検はさらに 59 個のアルファに及びました。**ユーザー報告からの修正:** 失敗した実行もステップを保持し、停止メッセージには何も得られなかった呼び出しが表示されます。`read_file` は読める場所を示すようになり、モデルが no-progress で止まるまでパスを推測し続けることはなくなりました。Codex では途中の指示がシステムプロンプトを置き換えなくなり、推論がターン間で引き継がれ、既定モデルは `gpt-6-sol` です。`get_market_data` は接尾辞のない米国株ティッカーに空データを返さず `AAPL.US` を求めます。**新機能:** アルゼンチン(BYMA)、週足と月足、インドネシア語、Gildata(恒生聚源)、OpenCode、メールと WebSocket チャネル、Web UI でのメッセージングチャネル設定、そして KIS、Upbit、Toss 証券、Scalable Capital のコネクタ(ブローカーは計 18 社)。[@Shizoqua](https://github.com/Shizoqua)、[@zeus229](https://github.com/zeus229)、[@cgycorey](https://github.com/cgycorey)、[@shadowinlife](https://github.com/shadowinlife)、[@lorenzozanee](https://github.com/lorenzozanee)、[@he-yufeng](https://github.com/he-yufeng)、[@chiww](https://github.com/chiww)、[@as950118](https://github.com/as950118)、[@woshi77777stars](https://github.com/woshi77777stars)、[@Yoruxyv](https://github.com/Yoruxyv)、[@sambazhu](https://github.com/sambazhu)、[@0xouzm](https://github.com/0xouzm)、[@tingkk](https://github.com/tingkk)、[@tonydo](https://github.com/tonydo)、[@alanwilhelm](https://github.com/alanwilhelm)、[@modelpath-dev](https://github.com/modelpath-dev) に感謝します!- **2026-09-28** 🛠️ **データソースの健全性と接続機能の可視化**:認証情報を使わず時間制限付きで公開ソースを定期検査し、接続障害やデータ異常を報告します([#1627](https://github.com/HKUDS/Vibe-Trading/pull/1627))。BaoStock に通信期限と並行セッションの直列化を追加([#1615](https://github.com/HKUDS/Vibe-Trading/pull/1615))、Copilot の認証情報照会には短いキャッシュ期限を設定しました([#1619](https://github.com/HKUDS/Vibe-Trading/pull/1619))。自動生成のブローカー機能表はペーパー/実取引の権限をプロファイル別に示し、宣言と動作確認を区別します([#1629](https://github.com/HKUDS/Vibe-Trading/pull/1629))。 **数値検証と MT5 の追加修正**:明示された銘柄と正確なフィールド参照を保持し、地域別の桁区切りにも既存の数値許容幅を適用します([#1584](https://github.com/HKUDS/Vibe-Trading/pull/1584)、[#1586](https://github.com/HKUDS/Vibe-Trading/pull/1586)、[#1588](https://github.com/HKUDS/Vibe-Trading/pull/1588))。テキストのみの修正回数を制限します([#1600](https://github.com/HKUDS/Vibe-Trading/pull/1600))。MT5 は指定端末で検索し、曖昧なブローカー別名を拒否します。バックテストのサンドボックスには検証済みの接続設定だけを渡します([#1597](https://github.com/HKUDS/Vibe-Trading/pull/1597)、[#1598](https://github.com/HKUDS/Vibe-Trading/pull/1598))。- **2026-09-27** 🛠️ **検証可能な実行カードと研究フローの改善**:JSON、Markdown、実行詳細に、ハッシュ化したバックテスト実行記録と検証済みの指標 CSV 参照を表示します([#1612](https://github.com/HKUDS/Vibe-Trading/pull/1612))。GTJA の高値・安値経過日数と同値時の処理を修正し([#1604](https://github.com/HKUDS/Vibe-Trading/pull/1604))、インド市場の空売り買い戻しは買い側の値幅制限を確認します([#1608](https://github.com/HKUDS/Vibe-Trading/pull/1608))。認証情報を露出せずトークン使用量を保持し([#1606](https://github.com/HKUDS/Vibe-Trading/pull/1606))、不正な MCP スキーマのオブジェクト属性を補完します([#1607](https://github.com/HKUDS/Vibe-Trading/pull/1607))。即時注文の目標チェックは研究目的の質問を許可し([#1605](https://github.com/HKUDS/Vibe-Trading/pull/1605))、設定文書の既定経路を OpenRouter と明記しました([#1609](https://github.com/HKUDS/Vibe-Trading/issues/1609))。+<details>+<summary>過去のニュース</summary>+- **2026-09-26** 🛠️ **指標計算と取引上限の信頼性を改善**:テクニカル指標は観測日を保持し、会話の圧縮で失われた結果を再利用できます([#1590](https://github.com/HKUDS/Vibe-Trading/pull/1590)、[#1601](https://github.com/HKUDS/Vibe-Trading/pull/1601))。Qlib158 の全5種類の WVMA 窓で分子に絶対リターンを使用し([#1594](https://github.com/HKUDS/Vibe-Trading/pull/1594))、Benford 検査は数値の境界でも正しい先頭桁を保持します([#1591](https://github.com/HKUDS/Vibe-Trading/pull/1591))。明示的にゼロに設定したエクスポージャー上限はゼロのまま維持され([#1593](https://github.com/HKUDS/Vibe-Trading/pull/1593))、Dhan の模擬注文は入力精度を丸めず、小数や無効な数量を拒否します([#1595](https://github.com/HKUDS/Vibe-Trading/pull/1595))。任意の数値環境変数が無効な場合はデフォルト値に戻り、明示的な設定の制約は引き続き適用されます([#1592](https://github.com/HKUDS/Vibe-Trading/pull/1592))。実行時トリガーのフィールド初期値とサブクラスのファクトリ動作も修正しました([#1599](https://github.com/HKUDS/Vibe-Trading/pull/1599))。- **2026-09-25** 🛠️ **リサーチセッションとチャネル診断を改善**:バックテスト CSV に文字列の列があっても数値指標が失われなくなり([#1579](https://github.com/HKUDS/Vibe-Trading/pull/1579))、会話圧縮で保持するメッセージの容量計算に推論内容を含めるようになりました([#1582](https://github.com/HKUDS/Vibe-Trading/pull/1582))。チャネルの読み込み失敗時には実際の例外を記録し([#1580](https://github.com/HKUDS/Vibe-Trading/pull/1580))、Telegram ガイドでは CLI と Web の操作が同じ API ランタイムを共有することを明記しました([#1583](https://github.com/HKUDS/Vibe-Trading/pull/1583))。Robinhood のポートフォリオ取得は不正な形式の購買力オブジェクトを拒否し、省略値や null は不明のまま保持します([#1526](https://github.com/HKUDS/Vibe-Trading/pull/1526))。チャット画面を離れると、履歴のスクロール待機タイマーも解除されます。[@Shizoqua](https://github.com/Shizoqua) と [@lorenzozanee](https://github.com/lorenzozanee) に感謝!-- **2026-09-24** 💱 **ペソで計上されていたドル建て銘柄、Web UI に加わった Email と WebSocket、そして SDK 更新による CI 障害**:BYMA とトロント証券取引所には自国通貨建てに加えて米ドル建ての銘柄(GGALD.BA、DLR-U.TO)があり、アルゼンチンやカナダのバックテストはそれらを単一のペソ/カナダドルのプールで評価していました。ローダーは市場の通貨で建値された銘柄だけを受け入れるようになり(LSE 銘柄と同じ規則)、それ以外に申告された建値通貨はデータとともに渡されるため、回答には正しい通貨が書かれます([#1576](https://github.com/HKUDS/Vibe-Trading/pull/1576))。香港取引所が独自のコード帯で採番する人民元建て・米ドル建てカウンター(80700.HK、9834.HK)も同じ理由で香港のバックテストから除外され、それぞれの通貨で扱われます。銘柄プロファイルは発行体のファンダメンタルズとは別に、その上場銘柄自身の取引所と通貨を返し([#1577](https://github.com/HKUDS/Vibe-Trading/pull/1577))、グラウンディングと銘柄検索はデータ層が扱うすべての市場を認識し([#1575](https://github.com/HKUDS/Vibe-Trading/pull/1575))、`technical_indicators` は出来高をデータソースが申告する単位(単元か株か、100 倍違う)付きで返します([#1571](https://github.com/HKUDS/Vibe-Trading/pull/1571))。Email と WebSocket がガイド付きチャネル設定に加わり、メールサーバー証明書の検証と、クロスサイトのページからの設定書き込みの拒否、SSL なしの IMAP でもパスワード送信前に STARTTLS で暗号化する変更も入りました([#1544](https://github.com/HKUDS/Vibe-Trading/pull/1544))。**修正:** openai 3.19.2 で明示したプロバイダーヘッダーが環境側の同名ヘッダーに負け、全 PR の CI が赤くなっていた問題([#1568](https://github.com/HKUDS/Vibe-Trading/pull/1568))、マイクロ秒精度のソースからのバックテストが 1970 年の日付になっていた問題([#1560](https://github.com/HKUDS/Vibe-Trading/pull/1560))、明らかに注文である研究ゴールを再び拒否([#1562](https://github.com/HKUDS/Vibe-Trading/pull/1562))、strict bench の OOS 分割で片側が空になりうる問題([#1559](https://github.com/HKUDS/Vibe-Trading/pull/1559))、HRP のラベル整列・交差検証のラベル順序・非有限な入力([#1555](https://github.com/HKUDS/Vibe-Trading/pull/1555)、[#1556](https://github.com/HKUDS/Vibe-Trading/pull/1556)、[#1557](https://github.com/HKUDS/Vibe-Trading/pull/1557)、[#1558](https://github.com/HKUDS/Vibe-Trading/pull/1558))、シャドーアカウントの過剰取引の期間([#1563](https://github.com/HKUDS/Vibe-Trading/pull/1563))、日次で比較されていた総資産成長率([#1564](https://github.com/HKUDS/Vibe-Trading/pull/1564))。[@Shizoqua](https://github.com/Shizoqua)、[@zeus229](https://github.com/zeus229)、[@shadowinlife](https://github.com/shadowinlife) に感謝します!-<details>-<summary>過去のニュース</summary>++- **2026-09-24** 💱 **ペソで計上されていたドル建て銘柄、Web UI に加わった Email と WebSocket、そして SDK 更新による CI 障害**:BYMA とトロント証券取引所には自国通貨建てに加えて米ドル建ての銘柄(GGALD.BA、DLR-U.TO)があり、アルゼンチンやカナダのバックテストはそれらを単一のペソ/カナダドルのプールで評価していました。ローダーは市場の通貨で建値された銘柄だけを受け入れるようになり(LSE 銘柄と同じ規則)、それ以外に申告された建値通貨はデータとともに渡されるため、回答には正しい通貨が書かれます([#1576](https://github.com/HKUDS/Vibe-Trading/pull/1576))。香港取引所が独自のコード帯で採番する人民元建て・米ドル建てカウンター(80700.HK、9834.HK)も同じ理由で香港のバックテストから除外され、それぞれの通貨で扱われます。銘柄プロファイルは発行体のファンダメンタルズとは別に、その上場銘柄自身の取引所と通貨を返し([#1577](https://github.com/HKUDS/Vibe-Trading/pull/1577))、グラウンディングと銘柄検索はデータ層が扱うすべての市場を認識し([#1575](https://github.com/HKUDS/Vibe-Trading/pull/1575))、`technical_indicators` は出来高をデータソースが申告する単位(単元か株か、100 倍違う)付きで返します([#1571](https://github.com/HKUDS/Vibe-Trading/pull/1571))。Email と WebSocket がガイド付きチャネル設定に加わり、メールサーバー証明書の検証と、クロスサイトのページからの設定書き込みの拒否、SSL なしの IMAP でもパスワード送信前に STARTTLS で暗号化する変更も入りました([#1544](https://github.com/HKUDS/Vibe-Trading/pull/1544))。**修正:** openai 3.19.2 で明示したプロバイダーヘッダーが環境側の同名ヘッダーに負け、全 PR の CI が赤くなっていた問題([#1568](https://github.com/HKUDS/Vibe-Trading/pull/1568))、マイクロ秒精度のソースからのバックテストが 1970 年の日付になっていた問題([#1560](https://github.com/HKUDS/Vibe-Trading/pull/1560))、明らかに注文である研究ゴールを再び拒否([#1562](https://github.com/HKUDS/Vibe-Trading/pull/1562))、strict bench の OOS 分割で片側が空になりうる問題([#1559](https://github.com/HKUDS/Vibe-Trading/pull/1559))、HRP のラベル整列・交差検証のラベル順序・非有限な入力([#1555](https://github.com/HKUDS/Vibe-Trading/pull/1555)、[#1556](https://github.com/HKUDS/Vibe-Trading/pull/1556)、[#1557](https://github.com/HKUDS/Vibe-Trading/pull/1557)、[#1558](https://github.com/HKUDS/Vibe-Trading/pull/1558))、シャドーアカウントの過剰取引の期間([#1563](https://github.com/HKUDS/Vibe-Trading/pull/1563))、日次で比較されていた総資産成長率([#1564](https://github.com/HKUDS/Vibe-Trading/pull/1564))。[@Shizoqua](https://github.com/Shizoqua)、[@zeus229](https://github.com/zeus229)、[@shadowinlife](https://github.com/shadowinlife) に感謝します!- **2026-09-23** 🇦🇷 **アルゼンチン銘柄、逆向きにサイズされていたショート、そしてテールリスク・ゲートの残り半分**:`.BA`(BYMA 株式と CEDEAR)は独自の ARS 市場として Yahoo に接続され、Web UI でも他の市場と同じように分類・表示されます。BYMA の執行ルールをモデル化するまで、アルゼンチンのバックテストは他市場の仮定を流用せず明示的に失敗します([#1543](https://github.com/HKUDS/Vibe-Trading/pull/1543))。平均分散型と回転率考慮型のオプティマイザはすべてのポジションをロングとして評価していたため、最良のショートに最も少ない資金しか配分されませんでした。どちらもポジション収益でサイズを決めるようになり、相関する 2 銘柄のロングとショートは相関ではなくヘッジとして読まれます(相関 +0.92 の組で、旧目的関数はロングに全額、新しい配分は 0.51 / -0.49)([#1548](https://github.com/HKUDS/Vibe-Trading/pull/1548))。さらに、1 つのセッションが複数のテールリスク指標(VaR と ES、95% と 99%)を保持した時点で、テールリスクの数値は自分のフィールドを明示しなければならず、ES 95% が VaR 95% の値を借りることはできません([#1425](https://github.com/HKUDS/Vibe-Trading/issues/1425))。**修正:**モンテカルロ検証は 252 固定ではなく取引所の年間バー数で年率化し、同じレポート内のシャープと矛盾しなくなりました([#1546](https://github.com/HKUDS/Vibe-Trading/pull/1546))。負値バーでの先物発注が一律に拒否されなくなりました([#1547](https://github.com/HKUDS/Vibe-Trading/pull/1547))。ベースライン IC が負の場合、ファクターの IC ratio は未設定になり、実際の劣化が健全な改善として読まれることはありません([#1549](https://github.com/HKUDS/Vibe-Trading/pull/1549))。タイトルが同じでタイプが異なる 2 つのメモリは、それぞれの索引行とリンク先を保ちます([#1545](https://github.com/HKUDS/Vibe-Trading/pull/1545))。[@zeus229](https://github.com/zeus229)、[@Shizoqua](https://github.com/Shizoqua)、[@he-yufeng](https://github.com/he-yufeng) に感謝します!diff --git a/README_ko.md b/README_ko.mdindex a64d6534e..18b448f92 100644--- a/README_ko.md+++ b/README_ko.md@@ -52,20 +52,22 @@> ⚠️ **보안 경고:** X 계정 `VibeTrading_HKU`, Virtuals 프로젝트 `101845`, 토큰 컨트랙트 `0x640BDBF77b6447E8b7DB7894cED84BD1c40571f4`는 모두 Vibe-Trading 공식과 무관합니다. Vibe-Trading은 어떠한 토큰이나 밈코인도 발행하거나 공식적으로 지지한 적이 없습니다. 해당 토큰을 구매하거나 지갑을 연결하거나 어떠한 서명도 하지 마세요. [자세히 보기](SECURITY.md#official-channels--impersonation).-- **2026-09-29** 🛠️ **긴 비교가 끝까지 완료되고, 백테스트 보고서의 수치가 유지됩니다**: 대화 압축이 고정된 40K 추정 대신 모델의 실제 컨텍스트 윈도우(Codex의 GPT-6은 272K, DeepSeek V4는 1M)에 맞춰 이루어집니다. 기존 기준은 세 번째 단계부터 도구 결과를 지워, 12개 기업 비교가 같은 재무 자료를 반복해서 다시 가져오다 no-progress로 중단되었습니다. 실패한 실행도 단계를 보존하며, 중단 메시지에 도구와 이유가 표시됩니다. Codex에서는 실행 중 지시가 시스템 프롬프트를 대체하지 않고, 추론이 턴 사이에 유지되며, 기본 모델은 `gpt-6-sol`입니다. 수치 검증이 백테스트 자체 출력(Sortino, Calmar, 회전율, 비중, 몬테카를로 p값)을 읽고 출처 실행을 밝힌 수치를 대조하므로, 리스크 패리티 대 동일가중 비교 보고서가 더 이상 거절 문구로 대체되지 않습니다. `get_market_data`는 접미사 없는 미국 주식 티커에 빈 데이터 대신 `AAPL.US`를 요청하고, 다중 시장 백테스트는 시장별로 데이터 소스를 대체합니다.+- **2026-09-29** 🚀 **v0.1.16 출시** ([릴리스 노트](https://github.com/HKUDS/Vibe-Trading/releases/tag/v0.1.16), `pip install -U vibe-trading-ai`): 0.1.15 이후 커밋 492개, 병합된 풀 리퀘스트 116개, 기여자 16명. **이번 사이클의 주제는 모든 숫자가 자신의 출처를 보여줄 수 있는 것입니다.** 그라운딩 게이트는 더 이상 숫자 주변의 단어로 그 의미를 추측하지 않습니다. 모델이 각 숫자의 역할(observed / derived / proposed / cited / count)을 선언하면 게이트가 해당 세션의 도구 결과와 대조하고, 검증되지 않은 숫자는 답변 전체를 거부하는 대신 그 숫자만 삭제합니다. 백테스트 자체의 출력(샤프 비율뿐 아니라 소르티노, 회전율, 비중, 몬테카를로 p값)이 그에 관한 보고서의 근거가 되며, 런 카드는 저장된 CSV와 일치하는 지표만 인용합니다. 긴 리서치 작업은 가져온 데이터를 더 이상 잃지 않습니다. 컨텍스트 압축이 고정된 40K 추정 대신 각 모델의 실제 컨텍스트 윈도우를 따릅니다. 결측 데이터 점검은 알파 59개로 더 확대되었습니다. **사용자 제보로 수정:** 실패한 실행도 단계를 보존하고 중단 메시지에 아무것도 얻지 못한 호출이 표시됩니다. `read_file`이 읽을 수 있는 위치를 알려 주어, 모델이 no-progress로 중단될 때까지 경로를 추측하지 않습니다. Codex에서는 실행 중 지시가 시스템 프롬프트를 대체하지 않고, 추론이 턴 사이에 유지되며, 기본 모델은 `gpt-6-sol`입니다. `get_market_data`는 접미사 없는 미국 주식 티커에 빈 데이터 대신 `AAPL.US`를 요청합니다. **신규:** 아르헨티나(BYMA), 주봉과 월봉, 인도네시아어, Gildata, OpenCode, 이메일과 WebSocket 채널, Web UI에서의 메신저 채널 설정, 그리고 한국투자증권(KIS), 업비트, 토스증권, Scalable Capital 커넥터(총 18개 브로커). [@Shizoqua](https://github.com/Shizoqua), [@zeus229](https://github.com/zeus229), [@cgycorey](https://github.com/cgycorey), [@shadowinlife](https://github.com/shadowinlife), [@lorenzozanee](https://github.com/lorenzozanee), [@he-yufeng](https://github.com/he-yufeng), [@chiww](https://github.com/chiww), [@as950118](https://github.com/as950118), [@woshi77777stars](https://github.com/woshi77777stars), [@Yoruxyv](https://github.com/Yoruxyv), [@sambazhu](https://github.com/sambazhu), [@0xouzm](https://github.com/0xouzm), [@tingkk](https://github.com/tingkk), [@tonydo](https://github.com/tonydo), [@alanwilhelm](https://github.com/alanwilhelm), [@modelpath-dev](https://github.com/modelpath-dev)에게 감사드립니다!- **2026-09-28** 🛠️ **데이터 소스 상태와 브로커 기능 공개**: 인증 정보 없이 시간 제한을 적용한 정기 검사로 공개 소스의 연결 장애와 데이터 이상을 보고합니다([#1627](https://github.com/HKUDS/Vibe-Trading/pull/1627)). BaoStock 통신 제한 시간과 동시 세션 직렬화를 추가하고([#1615](https://github.com/HKUDS/Vibe-Trading/pull/1615)), Copilot 자격 증명 조회 캐시에 짧은 만료 시간을 적용했습니다([#1619](https://github.com/HKUDS/Vibe-Trading/pull/1619)). 자동 생성 브로커 표는 프로필별 모의·실거래 권한을 구분하며, 선언된 기능과 실제 검증을 구별합니다([#1629](https://github.com/HKUDS/Vibe-Trading/pull/1629)). **수치 검증 및 MT5 추가 수정**: 명시된 종목과 정확한 필드 참조를 보존하고 현지 숫자 구분에도 기존 허용 오차를 유지합니다([#1584](https://github.com/HKUDS/Vibe-Trading/pull/1584), [#1586](https://github.com/HKUDS/Vibe-Trading/pull/1586), [#1588](https://github.com/HKUDS/Vibe-Trading/pull/1588)). 텍스트 전용 수정 횟수를 제한합니다([#1600](https://github.com/HKUDS/Vibe-Trading/pull/1600)). MT5는 지정 터미널에서 검색하며 모호한 브로커 별칭을 거부합니다. 백테스트 샌드박스에는 검증된 연결 설정만 전달합니다([#1597](https://github.com/HKUDS/Vibe-Trading/pull/1597), [#1598](https://github.com/HKUDS/Vibe-Trading/pull/1598)).- **2026-09-27** 🛠️ **검증 가능한 실행 카드와 연구 흐름 개선**: JSON, Markdown, 실행 상세 화면에 해시 기반 백테스트 실행 기록과 검증된 지표 CSV 참조를 표시합니다([#1612](https://github.com/HKUDS/Vibe-Trading/pull/1612)). GTJA 고점·저점 경과 일수와 동률 처리를 수정하고([#1604](https://github.com/HKUDS/Vibe-Trading/pull/1604)), 인도 시장 숏 청산은 매수 방향의 가격 제한을 확인합니다([#1608](https://github.com/HKUDS/Vibe-Trading/pull/1608)). 자격 증명 문자열을 노출하지 않고 토큰 사용량을 유지하며([#1606](https://github.com/HKUDS/Vibe-Trading/pull/1606)), 잘못된 MCP 스키마의 객체 속성을 복구합니다([#1607](https://github.com/HKUDS/Vibe-Trading/pull/1607)). 즉시 주문 목표 검사는 연구 질문을 허용하고([#1605](https://github.com/HKUDS/Vibe-Trading/pull/1605)), 설정 문서에 기본 경로가 OpenRouter임을 명시했습니다([#1609](https://github.com/HKUDS/Vibe-Trading/issues/1609)).+<details>+<summary>이전 뉴스</summary>+- **2026-09-26** 🛠️ **지표 계산과 거래 한도 개선**: 기술적 지표는 관측 날짜를 유지하고 대화 압축으로 사라진 결과를 재사용할 수 있습니다([#1590](https://github.com/HKUDS/Vibe-Trading/pull/1590), [#1601](https://github.com/HKUDS/Vibe-Trading/pull/1601)). Qlib158의 다섯 WVMA 기간 모두 분자에 절대 수익률을 사용하며([#1594](https://github.com/HKUDS/Vibe-Trading/pull/1594)), Benford 검사는 수치 경계에서도 올바른 첫 자릿수를 유지합니다([#1591](https://github.com/HKUDS/Vibe-Trading/pull/1591)). 명시적으로 0으로 설정한 익스포저 한도는 그대로 유지되고([#1593](https://github.com/HKUDS/Vibe-Trading/pull/1593)), Dhan 모의 주문은 입력 정밀도를 반올림으로 잃지 않으면서 소수 및 잘못된 수량을 거부합니다([#1595](https://github.com/HKUDS/Vibe-Trading/pull/1595)). 잘못된 선택적 숫자 환경 변수는 기본값으로 돌아가며 명시적 설정의 제약은 계속 적용됩니다([#1592](https://github.com/HKUDS/Vibe-Trading/pull/1592)). 런타임 트리거의 필드 기본값과 하위 클래스 팩터리 동작도 수정했습니다([#1599](https://github.com/HKUDS/Vibe-Trading/pull/1599)).- **2026-09-25** 🛠️ **리서치 세션과 채널 진단 개선**: 백테스트 CSV에 텍스트 열이 있어도 숫자 지표가 통째로 사라지지 않으며([#1579](https://github.com/HKUDS/Vibe-Trading/pull/1579)), 대화 압축 시 보존할 메시지의 용량 계산에 추론 내용도 포함합니다([#1582](https://github.com/HKUDS/Vibe-Trading/pull/1582)). 채널 로드 실패 시 실제 예외를 기록하고([#1580](https://github.com/HKUDS/Vibe-Trading/pull/1580)), Telegram 가이드에 CLI와 Web 제어가 같은 API 런타임을 공유한다는 점을 명시했습니다([#1583](https://github.com/HKUDS/Vibe-Trading/pull/1583)). Robinhood 포트폴리오 조회는 잘못된 형식의 구매력 객체를 거부하되, 누락되거나 null인 값은 알 수 없는 값으로 유지합니다([#1526](https://github.com/HKUDS/Vibe-Trading/pull/1526)). 채팅 화면을 떠나면 대기 중인 기록 스크롤 타이머도 취소합니다. [@Shizoqua](https://github.com/Shizoqua), [@lorenzozanee](https://github.com/lorenzozanee) 감사합니다!-- **2026-09-24** 💱 **페소로 계산되던 달러 종목, Web UI에 들어온 Email과 WebSocket, 그리고 SDK 업데이트로 깨진 CI**: BYMA와 토론토 증권거래소에는 자국 통화 종목 옆에 미국 달러 종목(GGALD.BA, DLR-U.TO)도 상장되어 있는데, 아르헨티나·캐나다 백테스트는 이를 하나뿐인 페소·캐나다 달러 풀에서 평가했습니다. 이제 로더는 해당 시장 통화로 호가되는 종목만 받아들이고(LSE 종목과 같은 규칙), 그 밖에 소스가 선언한 호가 통화는 데이터와 함께 전달되어 답변에 올바른 통화가 적힙니다([#1576](https://github.com/HKUDS/Vibe-Trading/pull/1576)). 홍콩거래소가 별도 코드 구간으로 번호를 매기는 위안화·달러 카운터(80700.HK, 9834.HK)도 같은 이유로 홍콩 백테스트에서 거부되고 각자의 통화로 표시됩니다. 종목 프로필은 발행사 펀더멘털과 별도로 해당 상장 종목의 거래소와 통화를 담고([#1577](https://github.com/HKUDS/Vibe-Trading/pull/1577)), 그라운딩과 종목 검색은 데이터 계층이 다루는 모든 시장을 인식하며([#1575](https://github.com/HKUDS/Vibe-Trading/pull/1575)), `technical_indicators`는 거래량을 소스가 선언한 단위(매매 단위 lot 또는 주, 100배 차이)와 함께 반환합니다([#1571](https://github.com/HKUDS/Vibe-Trading/pull/1571)). Email과 WebSocket이 가이드형 채널 설정에 합류했고, 메일 서버 인증서를 검증하며 크로스사이트 페이지의 설정 쓰기를 막으며, SSL 없는 IMAP 로그인도 비밀번호를 보내기 전에 STARTTLS로 암호화합니다([#1544](https://github.com/HKUDS/Vibe-Trading/pull/1544)). **수정:** openai 3.19.2에서 명시한 provider 헤더가 환경 변수의 같은 이름 헤더에 밀려 모든 PR의 CI가 빨갛게 되던 문제([#1568](https://github.com/HKUDS/Vibe-Trading/pull/1568)), 마이크로초 해상도 소스의 백테스트가 1970년으로 찍히던 문제([#1560](https://github.com/HKUDS/Vibe-Trading/pull/1560)), 명백한 주문인 리서치 목표를 다시 거부([#1562](https://github.com/HKUDS/Vibe-Trading/pull/1562)), strict bench의 OOS 분할이 한쪽을 비울 수 있던 문제([#1559](https://github.com/HKUDS/Vibe-Trading/pull/1559)), HRP 라벨 정렬·교차검증 라벨 순서·유한하지 않은 퀀트 입력([#1555](https://github.com/HKUDS/Vibe-Trading/pull/1555), [#1556](https://github.com/HKUDS/Vibe-Trading/pull/1556), [#1557](https://github.com/HKUDS/Vibe-Trading/pull/1557), [#1558](https://github.com/HKUDS/Vibe-Trading/pull/1558)), 섀도 계정의 과잉 매매 기간([#1563](https://github.com/HKUDS/Vibe-Trading/pull/1563)), 일 단위로 계산되던 자산 성장률([#1564](https://github.com/HKUDS/Vibe-Trading/pull/1564)). [@Shizoqua](https://github.com/Shizoqua), [@zeus229](https://github.com/zeus229), [@shadowinlife](https://github.com/shadowinlife) 님 감사합니다!-<details>-<summary>이전 뉴스</summary>++- **2026-09-24** 💱 **페소로 계산되던 달러 종목, Web UI에 들어온 Email과 WebSocket, 그리고 SDK 업데이트로 깨진 CI**: BYMA와 토론토 증권거래소에는 자국 통화 종목 옆에 미국 달러 종목(GGALD.BA, DLR-U.TO)도 상장되어 있는데, 아르헨티나·캐나다 백테스트는 이를 하나뿐인 페소·캐나다 달러 풀에서 평가했습니다. 이제 로더는 해당 시장 통화로 호가되는 종목만 받아들이고(LSE 종목과 같은 규칙), 그 밖에 소스가 선언한 호가 통화는 데이터와 함께 전달되어 답변에 올바른 통화가 적힙니다([#1576](https://github.com/HKUDS/Vibe-Trading/pull/1576)). 홍콩거래소가 별도 코드 구간으로 번호를 매기는 위안화·달러 카운터(80700.HK, 9834.HK)도 같은 이유로 홍콩 백테스트에서 거부되고 각자의 통화로 표시됩니다. 종목 프로필은 발행사 펀더멘털과 별도로 해당 상장 종목의 거래소와 통화를 담고([#1577](https://github.com/HKUDS/Vibe-Trading/pull/1577)), 그라운딩과 종목 검색은 데이터 계층이 다루는 모든 시장을 인식하며([#1575](https://github.com/HKUDS/Vibe-Trading/pull/1575)), `technical_indicators`는 거래량을 소스가 선언한 단위(매매 단위 lot 또는 주, 100배 차이)와 함께 반환합니다([#1571](https://github.com/HKUDS/Vibe-Trading/pull/1571)). Email과 WebSocket이 가이드형 채널 설정에 합류했고, 메일 서버 인증서를 검증하며 크로스사이트 페이지의 설정 쓰기를 막으며, SSL 없는 IMAP 로그인도 비밀번호를 보내기 전에 STARTTLS로 암호화합니다([#1544](https://github.com/HKUDS/Vibe-Trading/pull/1544)). **수정:** openai 3.19.2에서 명시한 provider 헤더가 환경 변수의 같은 이름 헤더에 밀려 모든 PR의 CI가 빨갛게 되던 문제([#1568](https://github.com/HKUDS/Vibe-Trading/pull/1568)), 마이크로초 해상도 소스의 백테스트가 1970년으로 찍히던 문제([#1560](https://github.com/HKUDS/Vibe-Trading/pull/1560)), 명백한 주문인 리서치 목표를 다시 거부([#1562](https://github.com/HKUDS/Vibe-Trading/pull/1562)), strict bench의 OOS 분할이 한쪽을 비울 수 있던 문제([#1559](https://github.com/HKUDS/Vibe-Trading/pull/1559)), HRP 라벨 정렬·교차검증 라벨 순서·유한하지 않은 퀀트 입력([#1555](https://github.com/HKUDS/Vibe-Trading/pull/1555), [#1556](https://github.com/HKUDS/Vibe-Trading/pull/1556), [#1557](https://github.com/HKUDS/Vibe-Trading/pull/1557), [#1558](https://github.com/HKUDS/Vibe-Trading/pull/1558)), 섀도 계정의 과잉 매매 기간([#1563](https://github.com/HKUDS/Vibe-Trading/pull/1563)), 일 단위로 계산되던 자산 성장률([#1564](https://github.com/HKUDS/Vibe-Trading/pull/1564)). [@Shizoqua](https://github.com/Shizoqua), [@zeus229](https://github.com/zeus229), [@shadowinlife](https://github.com/shadowinlife) 님 감사합니다!- **2026-09-23** 🇦🇷 **아르헨티나 종목, 거꾸로 사이징되던 숏 북, 그리고 테일리스크 게이트의 마지막 절반**: `.BA`(BYMA 주식과 CEDEAR)는 자체 ARS 시장으로 Yahoo에 연결되고 웹 UI도 다른 시장과 같이 묶어 표시합니다. BYMA 체결 규칙을 모델링하기 전까지 아르헨티나 백테스트는 다른 시장의 가정을 빌리지 않고 명시적으로 실패합니다([#1543](https://github.com/HKUDS/Vibe-Trading/pull/1543)). 평균-분산 및 회전율 고려 옵티마이저가 모든 포지션을 롱으로 평가해 가장 좋은 숏에 자본이 가장 적게 배분됐습니다. 이제 둘 다 포지션 수익으로 사이징하므로, 상관된 두 종목의 롱과 숏은 상관이 아니라 헤지로 읽힙니다(상관 +0.92인 쌍에서 기존 목적함수는 전액을 롱에, 새 방식은 0.51 / -0.49로 배분)([#1548](https://github.com/HKUDS/Vibe-Trading/pull/1548)). 또한 한 세션이 두 개 이상의 테일리스크 측정치(VaR와 ES, 95%와 99%)를 보유하면 테일리스크 수치는 자신의 필드를 지정해야 하며, ES 95%가 VaR 95% 값을 빌려 쓸 수 없습니다([#1425](https://github.com/HKUDS/Vibe-Trading/issues/1425)). **수정:** 몬테카를로 검증이 하드코딩된 252가 아니라 거래소의 연간 바 수로 연율화해 같은 리포트의 샤프와 어긋나지 않습니다([#1546](https://github.com/HKUDS/Vibe-Trading/pull/1546)). 음수 가격 바에서 선물 주문이 전부 거부되지 않습니다([#1547](https://github.com/HKUDS/Vibe-Trading/pull/1547)). 베이스라인 IC가 음수면 팩터의 IC ratio를 비워 두어 실제 감쇠를 건강한 개선으로 읽지 않습니다([#1549](https://github.com/HKUDS/Vibe-Trading/pull/1549)). 제목이 같고 타입이 다른 두 메모리는 각자의 색인 행과 링크 대상을 유지합니다([#1545](https://github.com/HKUDS/Vibe-Trading/pull/1545)). [@zeus229](https://github.com/zeus229), [@Shizoqua](https://github.com/Shizoqua), [@he-yufeng](https://github.com/he-yufeng) 님 감사합니다!diff --git a/README_zh.md b/README_zh.mdindex ab92c4424..a649b66b3 100644--- a/README_zh.md+++ b/README_zh.md@@ -52,20 +52,22 @@> ⚠️ **安全警告:** X 账号 `VibeTrading_HKU`、Virtuals 项目 `101845` 及代币合约 `0x640BDBF77b6447E8b7DB7894cED84BD1c40571f4` 均非 Vibe-Trading 官方。我们从未发行或背书任何代币或 meme 币。请勿购买、连接钱包或签名。[详细说明](SECURITY.md#official-channels--impersonation)。-- **2026-09-29** 🛠️ **长对比能跑完,回测报告不再被整篇拒答**:上下文压缩改按模型真实窗口计算(Codex 上的 GPT-6 为 272K,DeepSeek V4 为 1M),不再用固定的 40K 估算——旧阈值从第 3 步就开始清掉工具结果,12 家公司的对比会反复重取同一批财报,直到「无进展」停止。失败的运行保留步骤,停止说明写明是哪个工具、什么原因;Codex 下中途插入的指令不再覆盖系统提示,推理内容跨轮保留,默认模型改为 `gpt-6-sol`。数字核验现在能读到回测自己的输出(Sortino、Calmar、换手、权重、蒙特卡洛 p 值),数字写明来自哪次回测即可核对,风险平价与等权对比的报告不再被拒答替换。`get_market_data` 遇到不带后缀的美股代码会提示改用 `AAPL.US`,不再返回空数据;跨市场回测按各自市场回退数据源。+- **2026-09-29** 🚀 **v0.1.16 发布** ([发布说明](https://github.com/HKUDS/Vibe-Trading/releases/tag/v0.1.16), `pip install -U vibe-trading-ai`):自 0.1.15 以来共 492 个提交、116 个合并的 PR,来自 16 位贡献者。**这一版的主题是:每个数字都能说明自己从哪里来。** 数字核验不再根据数字旁边的文字去猜它是什么:模型为每个数字声明角色(observed / derived / proposed / cited / count),核验层对照本会话的工具结果逐个检查,核不上的数字单独删去,而不是整篇回答被拒。回测自己的输出——Sortino、换手、权重、蒙特卡洛 p 值,不只是夏普——现在都能用来核对关于它的报告;运行卡只在与保存的 CSV 一致时才引用某个指标。长研究任务不再丢失已取到的数据:上下文压缩按各模型的真实窗口计算,不再用固定的 40K 估算。缺失数据的清查又覆盖了 59 个因子。**本版还修复了用户反馈的问题:** 失败的运行保留步骤,停止说明写明是哪个调用没有产出;`read_file` 会说明能读哪里,模型不再反复猜路径直到「无进展」停止;Codex 下中途插入的指令不再覆盖系统提示,推理内容跨轮保留,默认模型改为 `gpt-6-sol`;`get_market_data` 遇到不带后缀的美股代码会提示改用 `AAPL.US`,不再返回空数据。**新增:** 阿根廷(BYMA)市场、周线和月线、印尼语界面、恒生聚源、OpenCode、邮件与 WebSocket 通道、在网页端配置即时通讯通道,以及韩国投资证券(KIS)、Upbit、Toss 证券和 Scalable Capital 连接器,券商总数达到 18 家。感谢 [@Shizoqua](https://github.com/Shizoqua), [@zeus229](https://github.com/zeus229), [@cgycorey](https://github.com/cgycorey), [@shadowinlife](https://github.com/shadowinlife), [@lorenzozanee](https://github.com/lorenzozanee), [@he-yufeng](https://github.com/he-yufeng), [@chiww](https://github.com/chiww), [@as950118](https://github.com/as950118), [@woshi77777stars](https://github.com/woshi77777stars), [@Yoruxyv](https://github.com/Yoruxyv), [@sambazhu](https://github.com/sambazhu), [@0xouzm](https://github.com/0xouzm), [@tingkk](https://github.com/tingkk), [@tonydo](https://github.com/tonydo), [@alanwilhelm](https://github.com/alanwilhelm) 和 [@modelpath-dev](https://github.com/modelpath-dev)!- **2026-09-28** 🛠️ **数据源健康检查与券商能力透明化**:定时检查通过有时限、不携带凭据的公开数据源探测报告连接故障和数据异常([#1627](https://github.com/HKUDS/Vibe-Trading/pull/1627));BaoStock socket 请求增加超时控制,并串行隔离并发会话([#1615](https://github.com/HKUDS/Vibe-Trading/pull/1615))。Copilot 凭据查询在短期缓存到期后重新读取([#1619](https://github.com/HKUDS/Vibe-Trading/pull/1619))。自动生成的券商矩阵分别列出各 profile 的模拟盘/实盘权限,明确区分能力声明与运行验证([#1629](https://github.com/HKUDS/Vibe-Trading/pull/1629))。 **数值核验与 MT5 补充修复**:保留证据中的显式标的及精确字段引用,本地化数字分组沿用既有数值容差([#1584](https://github.com/HKUDS/Vibe-Trading/pull/1584)、[#1586](https://github.com/HKUDS/Vibe-Trading/pull/1586)、[#1588](https://github.com/HKUDS/Vibe-Trading/pull/1588));纯文本纠错受次数限制([#1600](https://github.com/HKUDS/Vibe-Trading/pull/1600))。MT5 搜索使用指定终端,拒绝歧义券商后缀;回测沙箱只接收校验后的终端连接设置([#1597](https://github.com/HKUDS/Vibe-Trading/pull/1597)、[#1598](https://github.com/HKUDS/Vibe-Trading/pull/1598))。- **2026-09-27** 🛠️ **可复核的运行卡片与更可靠的研究流程**:运行卡片在 JSON、Markdown 和运行详情页展示回测执行哈希记录及已核验的指标 CSV 引用([#1612](https://github.com/HKUDS/Vibe-Trading/pull/1612))。GTJA 高低点远近因子修正计日与并列极值处理([#1604](https://github.com/HKUDS/Vibe-Trading/pull/1604)),印度市场回测平空仓按买入方向检查涨跌停([#1608](https://github.com/HKUDS/Vibe-Trading/pull/1608))。token 用量可读,同时不放行任意凭据字符串([#1606](https://github.com/HKUDS/Vibe-Trading/pull/1606));异常 MCP schema 保留对象属性([#1607](https://github.com/HKUDS/Vibe-Trading/pull/1607));即时下单目标检查不再误拦研究问题([#1605](https://github.com/HKUDS/Vibe-Trading/pull/1605))。配置文档明确默认使用 OpenRouter([#1609](https://github.com/HKUDS/Vibe-Trading/issues/1609))。+<details>+<summary>更早的新闻</summary>+- **2026-09-26** 🛠️ **指标计算与交易限额更可靠**:技术指标保留观测日期,并能恢复对话压缩时移除的结果,避免重复取数([#1590](https://github.com/HKUDS/Vibe-Trading/pull/1590)、[#1601](https://github.com/HKUDS/Vibe-Trading/pull/1601))。Qlib158 的五个 WVMA 窗口统一使用绝对收益率计算分子([#1594](https://github.com/HKUDS/Vibe-Trading/pull/1594));Benford 检查在数值边界也能保留正确首位数字([#1591](https://github.com/HKUDS/Vibe-Trading/pull/1591))。明确设为零的敞口上限保持为零([#1593](https://github.com/HKUDS/Vibe-Trading/pull/1593)),Dhan 模拟订单拒绝小数及无效数量,不再因舍入丢失输入精度([#1595](https://github.com/HKUDS/Vibe-Trading/pull/1595))。无效的可选数值环境变量回退到默认值,显式配置约束仍然生效([#1592](https://github.com/HKUDS/Vibe-Trading/pull/1592))。运行时触发器修复字段默认值,并保留子类工厂的正确行为([#1599](https://github.com/HKUDS/Vibe-Trading/pull/1599))。- **2026-09-25** 🛠️ **研究会话与频道诊断更可靠**:回测 CSV 中的文本列不再导致数值指标整行丢失([#1579](https://github.com/HKUDS/Vibe-Trading/pull/1579)),对话压缩的保留消息预算现在计入推理内容([#1582](https://github.com/HKUDS/Vibe-Trading/pull/1582))。频道加载失败时会正确记录异常([#1580](https://github.com/HKUDS/Vibe-Trading/pull/1580));Telegram 指南澄清了 CLI 与 Web 控件共享同一个 API 运行时([#1583](https://github.com/HKUDS/Vibe-Trading/pull/1583))。Robinhood 持仓读取拒绝格式错误的购买力对象,同时保留缺失或空值为未知([#1526](https://github.com/HKUDS/Vibe-Trading/pull/1526))。离开聊天页面时也会取消未执行的历史消息滚动定时器。感谢 [@Shizoqua](https://github.com/Shizoqua) 和 [@lorenzozanee](https://github.com/lorenzozanee)!-- **2026-09-24** 💱 **美元标的被按比索记账、Email 与 WebSocket 进入 Web UI,以及一次 SDK 升级引发的 CI 故障**:BYMA 和多伦多交易所在本币标的之外还挂有美元报价的标的(GGALD.BA、DLR-U.TO),阿根廷或加拿大回测却把它们放进唯一的比索或加元资金池计价。现在加载器只接纳以本市场货币报价的标的,与 LSE 标的的既有规则一致;其他由数据源声明的报价货币随数据一起传递,回答写出的就是正确的货币([#1576](https://github.com/HKUDS/Vibe-Trading/pull/1576))。港股的人民币柜台和美元柜台(80700.HK、9834.HK)在港交所有各自的代码区间,出于同样原因不再进入港股回测,并按其本身的货币报价。个股概况在发行人基本面之外给出该上市标的自己的交易所与货币([#1577](https://github.com/HKUDS/Vibe-Trading/pull/1577)),grounding 与标的搜索认识数据层支持的每个市场([#1575](https://github.com/HKUDS/Vibe-Trading/pull/1575)),`technical_indicators` 返回成交量时附上数据源声明的单位——手或股,相差 100 倍([#1571](https://github.com/HKUDS/Vibe-Trading/pull/1571))。Email 与 WebSocket 加入引导式频道配置,邮件服务器证书现在会被校验,设置写入也不再接受跨站页面的请求,不走 SSL 的 IMAP 登录也会先用 STARTTLS 加密再发送密码([#1544](https://github.com/HKUDS/Vibe-Trading/pull/1544))。**修复:** openai 3.19.2 下显式的 provider 请求头会被环境里的同名头挤掉,曾让所有 PR 的 CI 变红([#1568](https://github.com/HKUDS/Vibe-Trading/pull/1568));来自微秒精度数据源的回测被标成 1970 年([#1560](https://github.com/HKUDS/Vibe-Trading/pull/1560));明摆着是下单指令的研究目标重新被拒绝([#1562](https://github.com/HKUDS/Vibe-Trading/pull/1562));strict bench 的样本外切分可能让一侧为空([#1559](https://github.com/HKUDS/Vibe-Trading/pull/1559));HRP 标签对齐、交叉验证标签顺序和量化输入里的非有限值([#1555](https://github.com/HKUDS/Vibe-Trading/pull/1555)、[#1556](https://github.com/HKUDS/Vibe-Trading/pull/1556)、[#1557](https://github.com/HKUDS/Vibe-Trading/pull/1557)、[#1558](https://github.com/HKUDS/Vibe-Trading/pull/1558));影子账户过度交易的时间窗口([#1563](https://github.com/HKUDS/Vibe-Trading/pull/1563));以及按日而非按年计算的资产增长率([#1564](https://github.com/HKUDS/Vibe-Trading/pull/1564))。感谢 [@Shizoqua](https://github.com/Shizoqua)、[@zeus229](https://github.com/zeus229) 和 [@shadowinlife](https://github.com/shadowinlife)!-<details>-<summary>更早的新闻</summary>++- **2026-09-24** 💱 **美元标的被按比索记账、Email 与 WebSocket 进入 Web UI,以及一次 SDK 升级引发的 CI 故障**:BYMA 和多伦多交易所在本币标的之外还挂有美元报价的标的(GGALD.BA、DLR-U.TO),阿根廷或加拿大回测却把它们放进唯一的比索或加元资金池计价。现在加载器只接纳以本市场货币报价的标的,与 LSE 标的的既有规则一致;其他由数据源声明的报价货币随数据一起传递,回答写出的就是正确的货币([#1576](https://github.com/HKUDS/Vibe-Trading/pull/1576))。港股的人民币柜台和美元柜台(80700.HK、9834.HK)在港交所有各自的代码区间,出于同样原因不再进入港股回测,并按其本身的货币报价。个股概况在发行人基本面之外给出该上市标的自己的交易所与货币([#1577](https://github.com/HKUDS/Vibe-Trading/pull/1577)),grounding 与标的搜索认识数据层支持的每个市场([#1575](https://github.com/HKUDS/Vibe-Trading/pull/1575)),`technical_indicators` 返回成交量时附上数据源声明的单位——手或股,相差 100 倍([#1571](https://github.com/HKUDS/Vibe-Trading/pull/1571))。Email 与 WebSocket 加入引导式频道配置,邮件服务器证书现在会被校验,设置写入也不再接受跨站页面的请求,不走 SSL 的 IMAP 登录也会先用 STARTTLS 加密再发送密码([#1544](https://github.com/HKUDS/Vibe-Trading/pull/1544))。**修复:** openai 3.19.2 下显式的 provider 请求头会被环境里的同名头挤掉,曾让所有 PR 的 CI 变红([#1568](https://github.com/HKUDS/Vibe-Trading/pull/1568));来自微秒精度数据源的回测被标成 1970 年([#1560](https://github.com/HKUDS/Vibe-Trading/pull/1560));明摆着是下单指令的研究目标重新被拒绝([#1562](https://github.com/HKUDS/Vibe-Trading/pull/1562));strict bench 的样本外切分可能让一侧为空([#1559](https://github.com/HKUDS/Vibe-Trading/pull/1559));HRP 标签对齐、交叉验证标签顺序和量化输入里的非有限值([#1555](https://github.com/HKUDS/Vibe-Trading/pull/1555)、[#1556](https://github.com/HKUDS/Vibe-Trading/pull/1556)、[#1557](https://github.com/HKUDS/Vibe-Trading/pull/1557)、[#1558](https://github.com/HKUDS/Vibe-Trading/pull/1558));影子账户过度交易的时间窗口([#1563](https://github.com/HKUDS/Vibe-Trading/pull/1563));以及按日而非按年计算的资产增长率([#1564](https://github.com/HKUDS/Vibe-Trading/pull/1564))。感谢 [@Shizoqua](https://github.com/Shizoqua)、[@zeus229](https://github.com/zeus229) 和 [@shadowinlife](https://github.com/shadowinlife)!- **2026-09-23** 🇦🇷 **阿根廷标的、方向反了的空头仓位定权,以及尾部风险闸门的最后一半**:`.BA`(BYMA 股票与 CEDEAR)作为独立的 ARS 市场走 Yahoo,Web UI 也按普通市场分组和显示;阿根廷回测在 BYMA 执行规则建模之前仍然直接拒绝,不套用别的市场假设([#1543](https://github.com/HKUDS/Vibe-Trading/pull/1543))。均值方差和考虑换手的优化器把每个仓位都按做多打分,于是最好的空头拿到最少资金。两者现在按仓位收益定权:两只正相关标的一多一空是互相对冲,而不是读成同向相关——在相关性 +0.92 的一对上,旧目标函数把全部资金压在多头,新的分成 0.51 / -0.49([#1548](https://github.com/HKUDS/Vibe-Trading/pull/1548))。另外,一旦一次会话里出现多于一个尾部风险口径(VaR 与 ES,或 95% 与 99%),尾部风险数字必须指明自己的字段,ES 95% 再也不能借用 VaR 95% 的数值([#1425](https://github.com/HKUDS/Vibe-Trading/issues/1425))。**修复:**蒙特卡洛校验按交易场所的年 bar 数折算,而不是硬编码 252,同一份报告里的夏普不再自相矛盾([#1546](https://github.com/HKUDS/Vibe-Trading/pull/1546));负价 bar 上的期货下单不再被整单拒绝([#1547](https://github.com/HKUDS/Vibe-Trading/pull/1547));基准 IC 为负时因子的 IC ratio 留空,不再把真实衰减读成健康改善([#1549](https://github.com/HKUDS/Vibe-Trading/pull/1549));同名但类型不同的两条记忆各自保留索引行和链接目标([#1545](https://github.com/HKUDS/Vibe-Trading/pull/1545))。感谢 [@zeus229](https://github.com/zeus229)、[@Shizoqua](https://github.com/Shizoqua) 和 [@he-yufeng](https://github.com/he-yufeng)!diff --git a/agent/SKILL.md b/agent/SKILL.mdindex 987b07fc4..225015f56 100644--- a/agent/SKILL.md+++ b/agent/SKILL.md@@ -1,6 +1,6 @@---name: vibe-trading-version: 0.1.15+version: 0.1.16description: Professional finance research toolkit — backtesting (10 engines + benchmark comparison panel), factor analysis, Alpha Zoo (462 pre-built alphas across qlib158/alpha101/gtja191/academic/fundamental), options pricing, 90 finance skills, 30 multi-agent swarm teams, Trade Journal analyzer, and Shadow Account (extract → backtest → render) across 28 market-data sources (tushare, yfinance, okx, binance, akshare, baostock, tencent, mootdx, ccxt, futu, mt5, tickerall, local, eastmoney, sina, stooq, yahoo, pykrx, india_broker, qveris, longbridge, nobitex, wallex, plus optional-key finnhub/alphavantage/tiingo/fmp/gildata).dependencies:python: ">=3.11"diff --git a/frontend/package-lock.json b/frontend/package-lock.jsonindex 1801b6a63..25d9165cc 100644--- a/frontend/package-lock.json+++ b/frontend/package-lock.json@@ -1,12 +1,12 @@{"name": "vibe-trading-frontend",- "version": "0.1.15",+ "version": "0.1.16","lockfileVersion": 3,"requires": true,"packages": {"": {"name": "vibe-trading-frontend",- "version": "0.1.15",+ "version": "0.1.16","dependencies": {"@fontsource/inter": "^5.3.0","@fontsource/jetbrains-mono": "^5.3.0",diff --git a/frontend/package.json b/frontend/package.jsonindex c54c6a414..ecdcb5dcf 100644--- a/frontend/package.json+++ b/frontend/package.json@@ -1,7 +1,7 @@{"name": "vibe-trading-frontend","private": true,- "version": "0.1.15",+ "version": "0.1.16","type": "module","engines": {"node": ">=22.22.0"diff --git a/frontend/src/components/layout/__tests__/Layout.test.tsx b/frontend/src/components/layout/__tests__/Layout.test.tsxindex f0208456a..981011a14 100644--- a/frontend/src/components/layout/__tests__/Layout.test.tsx+++ b/frontend/src/components/layout/__tests__/Layout.test.tsx@@ -12,7 +12,7 @@ const sessions = [vi.mock("react-i18next", () => ({useTranslation: () => ({t: (key: string) => ({- "app.version": "v0.1.15",+ "app.version": "v0.1.16","layout.agent": "Agent","layout.alphaZoo": "Alpha Zoo","layout.cancel": "Cancel",diff --git a/frontend/src/i18n/locales/ar.json b/frontend/src/i18n/locales/ar.jsonindex 9a5237bb8..2c887bc5e 100644--- a/frontend/src/i18n/locales/ar.json+++ b/frontend/src/i18n/locales/ar.json@@ -1,6 +1,6 @@{"app": {- "version": "v0.1.15"+ "version": "v0.1.16"},"layout": {"home": "الرئيسية",diff --git a/frontend/src/i18n/locales/de.json b/frontend/src/i18n/locales/de.jsonindex f6b2b5d8a..2f751a095 100644--- a/frontend/src/i18n/locales/de.json+++ b/frontend/src/i18n/locales/de.json@@ -1,6 +1,6 @@{"app": {- "version": "v0.1.15"+ "version": "v0.1.16"},"charts": {"noPriceData": "Noch keine Kursdaten für dieses Symbol",diff --git a/frontend/src/i18n/locales/en.json b/frontend/src/i18n/locales/en.jsonindex 1b1cee407..0fc10bae8 100644--- a/frontend/src/i18n/locales/en.json+++ b/frontend/src/i18n/locales/en.json@@ -1,6 +1,6 @@{"app": {- "version": "v0.1.15"+ "version": "v0.1.16"},"charts": {"noPriceData": "No price data for this symbol yet",diff --git a/frontend/src/i18n/locales/es.json b/frontend/src/i18n/locales/es.jsonindex 887188d0e..d10e9da86 100644--- a/frontend/src/i18n/locales/es.json+++ b/frontend/src/i18n/locales/es.json@@ -1,6 +1,6 @@{"app": {- "version": "v0.1.15"+ "version": "v0.1.16"},"charts": {"noPriceData": "Aún no hay datos de precio para este símbolo",diff --git a/frontend/src/i18n/locales/id.json b/frontend/src/i18n/locales/id.jsonindex 0a0f125fe..bccdc71eb 100644--- a/frontend/src/i18n/locales/id.json+++ b/frontend/src/i18n/locales/id.json@@ -1,6 +1,6 @@{"app": {- "version": "v0.1.15"+ "version": "v0.1.16"},"charts": {"noPriceData": "Belum ada data harga untuk simbol ini",diff --git a/frontend/src/i18n/locales/ja.json b/frontend/src/i18n/locales/ja.jsonindex 50e88a1d0..bc000a218 100644--- a/frontend/src/i18n/locales/ja.json+++ b/frontend/src/i18n/locales/ja.json@@ -1,6 +1,6 @@{"app": {- "version": "v0.1.15"+ "version": "v0.1.16"},"layout": {"home": "ホーム",diff --git a/frontend/src/i18n/locales/ko.json b/frontend/src/i18n/locales/ko.jsonindex 1f3aa4aa2..d39ad72bc 100644--- a/frontend/src/i18n/locales/ko.json+++ b/frontend/src/i18n/locales/ko.json@@ -1,6 +1,6 @@{"app": {- "version": "v0.1.15"+ "version": "v0.1.16"},"layout": {"home": "홈",diff --git a/frontend/src/i18n/locales/pt-BR.json b/frontend/src/i18n/locales/pt-BR.jsonindex 7706459b8..c855c7fe9 100644--- a/frontend/src/i18n/locales/pt-BR.json+++ b/frontend/src/i18n/locales/pt-BR.json@@ -1,6 +1,6 @@{"app": {- "version": "v0.1.15"+ "version": "v0.1.16"},"charts": {"noPriceData": "Ainda não há dados de preço para este símbolo",diff --git a/frontend/src/i18n/locales/zh-CN.json b/frontend/src/i18n/locales/zh-CN.jsonindex 3e8f5481e..aabbe46b0 100644--- a/frontend/src/i18n/locales/zh-CN.json+++ b/frontend/src/i18n/locales/zh-CN.json@@ -1,6 +1,6 @@{"app": {- "version": "v0.1.15"+ "version": "v0.1.16"},"layout": {"home": "首页",diff --git a/pyproject.toml b/pyproject.tomlindex 37d6d5edd..4169d45bb 100644--- a/pyproject.toml+++ b/pyproject.toml@@ -1,6 +1,6 @@[project]name = "vibe-trading-ai"-version = "0.1.15"+version = "0.1.16"description = "Natural-language finance research AI agent with backtesting"requires-python = ">=3.11"license = "MIT"