Skip to content

Open-source platform for 3D AI agents. Turn text or a photo into a rigged, animated GLB avatar, give it an LLM brain, memory and a wallet, and embed it anywhere with one web component. Remote MCP server, x402 payments, Three.js, browser-native.

docs(changelog): publish the news-image and rivalries entries and their pages

nirholas023a95c554d1
diff --git a/CHANGELOG.md b/CHANGELOG.md
index 648a128f4..33df81d10 100644
--- a/CHANGELOG.md
+++ b/CHANGELOG.md
@@ -17,8 +17,10 @@ Public history for [three.ws](https://three.ws), newest first. New pages come fr
- **Endpoint Shopper fits a phone, and says something useful when a run finds nothing to buy**: A full audit of /shopper, the page where you describe a task, set a budget, and let an agent shop the x402 Bazaar for endpoints that can answer it. Three things were wrong. A run that came back with no plan and no answer, which is what happens when nothing in the Bazaar matches the task or everything that does prices above your ceiling, painted an empty white column and stopped: no explanation, no way forward. That case now says no endpoints matched, quotes the task back, and offers the three things that change the outcome (rewrite the task, raise the budget, browse the Bazaar) plus a button to run again. Second, when the request could not reach us at all, the page showed the browser's own wording for a dropped connection, which names no cause and suggests no fix; it now says the agent service could not be reached and what to do about it. Third, on a phone narrower than 560px the four navigation links were squeezed into a strip about 76px wide and had to be dragged sideways a few pixels at a time to read; they now wrap onto their own full-width row. Alongside those: the results column announces itself to a screen reader while a run is in flight and when it settles, the Task label sits above its box as intended rather than beside it, and an internal ticket number that had been sitting in the page's headline label since launch is gone. (`/shopper`) `[fix, improvement]`
- **Ethereum wallets load on the portfolio page again, keylessly, and every token logo shows up**: A full audit of /portfolio, where you paste any Solana or Ethereum address and get a live portfolio. The Ethereum half had gone dark: the single provider behind it ran out of monthly capacity, so every 0x address answered "Data source unavailable" with nothing behind it. Ethereum now has the same keyless floor Solana has always had. The ETH balance comes from the public Ethereum nodes the rest of the platform already uses, the token list with its prices and icons comes from a public block explorer, and the 24h move comes from the same market source the Solana side uses, with the coverage it achieved stated on the card. The Data panel names whichever lane actually answered, so you can see whether you are reading the paid provider or the free one. Token logos were another visible break: the page pointed your browser straight at an IPFS gateway, which the browser refuses on cross-origin grounds, so every holding fell back to a three-letter monogram and your address bar handed your IP to the gateway on every load. Logos now come through three.ws, sized for the row. Keyboard users could tab onto a slice of the allocation donut and get nothing: the tooltip was wired to an event handler property that browsers do not actually support, so it never fired once; the slice now names its asset and share, and Escape dismisses it. When a lookup fails, the page now shows the reason the server actually sent instead of a generic sentence, and clicking Retry after a rejected address no longer silently reloads the wallet you looked at before it. Underneath, the fallback that names Solana tokens when the primary metadata source is out of quota had never run at all, because of a missing import; token symbols and names come back on that path. (`/portfolio`) `[fix, improvement]`
- **Fact-check now reads the part of a source that is actually about your claim**: When the fact-checker fell back to Wikipedia, it only ever read each article's opening summary. An opening summary says what a page is about, so it carries the answer only when your claim is about the page's subject. Ask whether Napoleon was unusually short and it read the first lines of "Napoleon", "Napoleon III" and "Napoleonic Wars", none of which mention height at all, so every source was scored as saying nothing and the verdict came back "not enough evidence" while the passage about his height sat further down the same article. The checker now reads each article in full and picks the paragraphs that actually discuss your claim, so that question returns the "Appearance and image" and "Napoleon's height" sections instead. Reference lists are skipped, since they repeat your words without asserting anything, and the deeper read is bounded by the time the check already had, so a slow source can never cost you the evidence that already arrived. (`/docs/fact-check`) `[improvement]`
+- **Fact-check now reads the part of a source that is actually about your claim**: When the fact-checker fell back to Wikipedia, it only ever read each article's opening summary. An opening summary says what a page is about, so it carries the answer only when your claim is about the page's subject. Ask whether Napoleon was unusually short and it read the first lines of "Napoleon", "Napoleon III" and "Napoleonic Wars", none of which mention height at all, so every source was scored as saying nothing and the verdict came back "not enough evidence" while the passage about his height sat further down the same article. The checker now reads each article in full and picks the paragraphs that actually discuss your claim, so that question returns the "Appearance and image" and "Napoleon's height" sections instead. Reference lists are skipped, since they repeat your words without asserting anything, and the deeper read is bounded by the time the check already had, so a slow source can never cost you the evidence that already arrived. (`/docs/fact-check`) `[improvement]`
- **Fade Radar: the wallets that have never once been right, and the coins they are crowding into**: The Smart Money Radar tells you which proven winners are buying a coin. Fade Radar is the other half of that question, and on pump.fun it turns out to be the sharper one: how much of a coin's buy side is money that has never been right. A wallet joins the reverse-indicator board only after five coins it bought have a known outcome and not one of them won, which is a bar a surprising number of addresses clear. The new page ranks every launch from the last few hours by how much of its observed buy side belongs to those wallets, by wallet count and by SOL, and carries a second board of the wallets themselves with their full record. The number above the board is the part we care most about: it is measured, not asserted. We rebuild the reverse-indicator group using only coins whose outcome was known before the coins it is graded on existed, then check what actually happened. Coins where a quarter or more of the buyers were reverse indicators went on to win 1.8 percent of the time, against 22.1 percent for coins with none, across 26,626 labelled coins. Nothing here is a short and nothing signs anything: on a bonding curve the only tradeable form of a fade is not buying, so every action on the page is a link out to the coin, its intel, or a wallet's track record. Wallet links also now open directly on the Smart Money Radar, so an address cited anywhere on the platform lands on its record instead of a board you have to scan by eye. (`/fade`) `[feature]`
- **Homes connected through the relay now follow the house when you rearrange it**: If your Home Assistant is only reachable inside your house, three.ws reaches it through the three.ws integration you install there, which dials out to us. Those homes were building their room list once, when the connection opened, and then quietly ignoring changes: assign a lamp to the kitchen, rename a floor, add a device, and the 3D house kept showing the old arrangement until the connection happened to be rebuilt. Homes connected by a direct address never had this problem, and the split was in the relay, which was not carrying the four small notices Home Assistant sends when its rooms, floors, devices or entities change. It carries them now, so both kinds of connection track the house the same way and updates land in seconds. The notices only say that something changed; the relay already read the full room and floor lists, so nothing new about your house is exposed by this. (`/smart-home`) `[fix, infra]`
+- **News cards stop losing their preview picture to a cold cache**: About a fifth of the crypto-news feed ships text-only, so those cards start as a source tile and upgrade in place once the publisher's own preview image is found. That lookup searched all 197 publisher feeds at once and gave up after 2.5 seconds, which on a freshly started server dropped whichever feed had not answered yet, so an article the feed had just served came back as unknown. The reader kept its tile, but the failed lookup was recorded as a public answer for the next ten minutes, so every visitor to /markets and /markets/news got the same dead result and a 404 in their browser console. The lookup now asks the one publisher that could hold the article first and waits for it, and a result that could not be confirmed is held for 30 seconds instead of ten minutes. (`/markets/news`) `[fix]`
- **Oracle: a failed request now says so, instead of claiming no coins have been scored**: A full audit of /oracle, the fused conviction board for pump.fun launches. When a panel's request failed, the page told you the opposite of the truth: the live feed claimed "Oracle is warming up" and said the engine ships with its backend, the movers, proof, agent-ranking, activity and 3D-graph panels each claimed there was simply nothing to show, and the performance record said the edge was still proving itself. Every one of those is now a real error state that names what went wrong (offline, rate limited, or an engine error) and offers a retry that actually recovers. The Proof gallery had a second bug behind that one: opening it for the first time while the request failed painted nothing at all, not even a skeleton. Searching a symbol Oracle has not scored used to make the results dropdown vanish with no explanation, which reads as a broken search box; it now says there is no match, and says so differently when search itself is down. The conviction-breadth bar's tier segments and legend chips were 10 and 15 pixels tall, too small to reliably tap on a phone, and now carry a full 24-pixel target without changing how the bar looks. Screen readers can finally move from a tab into the panel it opens, every button declares its type, and if a browser is holding a cached copy of an older release the 3D graph explains that and offers a reload instead of retrying a module that can never load. (`/oracle`) `[fix, improvement]`
- **Paid 3D services now refuse a job they cannot deliver, instead of charging for it**: The paid three.ws forge services sold to other AI agents charge the moment the generator accepts a job, and the finished model reaches the buyer only after it lands in our delivery storage. When that storage stopped accepting writes, jobs were still being accepted and charged while no model could ever be handed back. Those services now check that delivery storage is writable before they take the job: if it is not, the call is refused, says so, and is not charged. The check is cached and ignores a one-off blip, so a passing wobble never stops the services selling. The free public health page also had a bug where a storage failure replaced the name of the failing subsystem with the storage provider's error code and printed the request's signing details; it now names the subsystem and reports the reason alone. `[fix, security]`
- **Pay-per-call is keyboard-drivable, and it now tells you when the wallet lane is down instead of sitting on a spinner**: A full audit of /pay, the page where an agent pays $0.001 USDC on Solana per call to a paid API. Three things were quietly broken. The wallet pill kept reverting to "loading" after it had already shown a real balance: the page fills that pill from a live balance lookup, but the translation pass that runs a moment later put the placeholder back over it, and the same race hit the session ticker and the wallet setup dialog. Those slots are now marked as owned by the page, so a rendered balance stays put. The whole wallet picker was unreachable without a mouse, because each wallet was a clickable box rather than a button; every wallet is now a real button that tab reaches, Enter and Space activate, and a screen reader announces as selected or not. Both dialogs (the first-visit walkthrough and the wallet setup prompt) now take focus when they open, close on Escape, hand focus back to whatever opened them, and stay hidden from screen readers while closed. States are honest now too. When the agent wallet service is unreachable the page used to tell a signed-out visitor to create their first agent, which was never the problem; it now says wallets are unavailable and offers a Retry that re-runs the lookup without a page reload. A failed activity feed used to leave a row of dots ticking forever and now says the feed is unavailable. The page also gained the heading it never had, one high-contrast focus ring across every control, and a copy pass that removed the long dashes from its description and on-page text. (`/pay`) `[fix, improvement]`
diff --git a/STRUCTURE.md b/STRUCTURE.md
index 109077169..9108da488 100644
--- a/STRUCTURE.md
+++ b/STRUCTURE.md
@@ -78,7 +78,7 @@ bottom.
| Skill bundles (seller builder + evidence pricing) | [pages/bundles.html](pages/bundles.html) `→ /bundles`, API [api/agents/[id]/bundles.js](api/agents/%5Bid%5D/bundles.js), arithmetic [api/_lib/bundle-pricing.js](api/_lib/bundle-pricing.js), checkout [api/marketplace/purchase-bundle.js](api/marketplace/purchase-bundle.js) | Live (seller side); no buyer-facing browser view yet | Sell several of one agent's skills as one purchase. The price is derived from the agent's own ledger rather than a rule of thumb: the public `?action=pricing` endpoint finds every buyer who already bought 2+ of the selected skills, suggests the median of what they really paid (`basis: median_basket`), and backtests any candidate price against those same baskets; with no such buyer yet it falls back to 20% off the sum of the parts and labels itself `discount_off_list` so a default is never read as a finding. Every amount is an atomic integer carried as a BigInt end to end, because a 9-decimal mint puts real sums past `Number.MAX_SAFE_INTEGER`. Checkout settles on-chain and writes one `skill_purchases` row per skill (`kind = 'bundle'`, amount 0) so one bundle sale is never counted as N skill sales. Docs: [docs/skill-bundles.md](docs/skill-bundles.md). Tests: [tests/bundle-pricing.test.js](tests/bundle-pricing.test.js), [tests/vercel-agents-subpath-routes.test.js](tests/vercel-agents-subpath-routes.test.js) |
| Marketplace analytics & trial funnel | [pages/marketplace-analytics.html](pages/marketplace-analytics.html) `→ /marketplace/analytics`, renderer [src/marketplace-analytics.js](src/marketplace-analytics.js), backend [api/marketplace/analytics.js](api/marketplace/analytics.js) | Live | Public transparency page for the skill marketplace: paid sales, volume, top skills/agents, and the trial conversion funnel. **Every money aggregate gates on `status = 'confirmed'` alone.** Counting `status IN ('confirmed','trial')` as a sale published 10,454 "sales" and 6.2M $THREE of "volume" on 2026-07-30 against a table holding zero confirmed purchases, because a `trial` row is a free grant whose `amount` is only the list price it would have cost. Free trials now render as their own stat and as a four-stage funnel (granted → used → exhausted → converted), which also names the two shapes that mean a wiring fault rather than weak demand: trials never spent, or trials exhausted with nothing converting. Docs: [docs/marketplace.md](docs/marketplace.md), tutorial [docs/tutorials/sell-a-skill-with-a-trial.md](docs/tutorials/sell-a-skill-with-a-trial.md). Tests: [tests/marketplace-analytics-paid-only.test.js](tests/marketplace-analytics-paid-only.test.js) |
| OKX.AI marketplace services | [api/okx/3d/[service].js](api/okx/3d/%5Bservice%5D.js) `→ /api/okx/3d/*` | Live | 3D services sold to other agents on OKX.AI (agent #2632). The **listed** line-up (rebuilt 2026-08-22) is **three.ws Forge**: four paid A2MCP rows (`forge-draft` $0.01 / `forge-standard` $0.05 / `forge-hd` $0.25 / `forge-image` $0.25) plus a free `forge-status` poll, tools in [api/_okx3d/forge.js](api/_okx3d/forge.js). Each row is a real MCP Streamable HTTP server whose paid `forge_3d` tool 402s with `eip155:196` leading; the buyer receives exactly what the ChatGPT custom GPT receives (GLB + concept image + viewer + AR link), because both fronts shape responses through [api/_mcp-studio/studio-shape.js](api/_mcp-studio/studio-shape.js). Agent Identity Studio and the single-capability REST endpoints are on the **back burner**: `listed: false` keeps them out of the listing and the submission payload while they stay deployed and payable. Catalog source of truth: [api/_lib/okx-catalog.js](api/_lib/okx-catalog.js); pipeline: [api/_okx3d/](api/_okx3d); showcase: [pages/agent-identities.html](pages/agent-identities.html) `→ /agent-identities` + [src/agent-identities.js](src/agent-identities.js), fed at runtime by [api/agent-identities.js](api/agent-identities.js) `→ /api/agent-identities` (demo runs from `data/agent-identities.json` merged with the live catalog row, so the price the page quotes is the price the paid endpoint charges); docs: [docs/okx-marketplace.md](docs/okx-marketplace.md), [docs/agent-identities.md](docs/agent-identities.md). Tests: [tests/api/okx-forge.test.js](tests/api/okx-forge.test.js) |
-| OKX.AI marketplace chat bot host | [workers/okx-chat-bot/](workers/okx-chat-bot) (Cloud Run service `okx-chat-bot`), local revive [scripts/okx-bot-revive.mjs](scripts/okx-bot-revive.mjs) `→ npm run okx:bot`, session seeding [scripts/okx-bot-seed-state.mjs](scripts/okx-bot-seed-state.mjs) `→ npm run okx:bot:seed-state` | Built · deploy owner-gated | Always-on host for the XMTP chat behind agent #2632. Supervises the `okx-a2a` daemon plus the `onchainos` wallet session, persists the wallet/XMTP identity to GCS across revisions, and rebuilds the AI subsession's briefing from [api/_lib/okx-chat-briefing.js](api/_lib/okx-chat-briefing.js) on every boot so quoted prices can never drift from the catalog. It exists because the stack used to live on a codespace that sleeps, and its outage mode is silent: an expired wallet session leaves the process healthy while no chat is delivered, which flagged the listing offline once. Health reaches the platform as the `okx_chat_bot` subsystem on `/api/healthz` via a `bot_heartbeat` row; a logged-out session pages with the exact login commands. The AI credential is probed against the provider rather than assumed present ([workers/okx-chat-bot/provider.js](workers/okx-chat-bot/provider.js)), because a key the provider refuses is indistinguishable from a working one until a buyer's message goes unanswered. Vertex AI is the deploy's transport, so the runtime service account is the credential and there is no secret to mint. Deploy: [workers/okx-chat-bot/cloudbuild.yaml](workers/okx-chat-bot/cloudbuild.yaml) (`--max-instances=1` is load-bearing, one GCS state writer). Tests: [tests/okx-chat-bot.test.js](tests/okx-chat-bot.test.js) |
+| OKX.AI marketplace chat bot host | [workers/okx-chat-bot/](workers/okx-chat-bot) (Cloud Run service `okx-chat-bot`), local revive [scripts/okx-bot-revive.mjs](scripts/okx-bot-revive.mjs) `→ npm run okx:bot`, session seeding [scripts/okx-bot-seed-state.mjs](scripts/okx-bot-seed-state.mjs) `→ npm run okx:bot:seed-state` | Live · Cloud Run since 2026-09-04, replies blocked on AI funding | Always-on host for the XMTP chat behind agent #2632. Supervises the `okx-a2a` daemon plus the `onchainos` wallet session, persists the wallet/XMTP identity to GCS across revisions, and rebuilds the AI subsession's briefing from [api/_lib/okx-chat-briefing.js](api/_lib/okx-chat-briefing.js) on every boot so quoted prices can never drift from the catalog. It exists because the stack used to live on a codespace that sleeps, and its outage mode is silent: an expired wallet session leaves the process healthy while no chat is delivered, which flagged the listing offline once. Health reaches the platform as the `okx_chat_bot` subsystem on `/api/healthz` via a `bot_heartbeat` row; a logged-out session pages with the exact login commands. The AI credential is probed against the provider rather than assumed present ([workers/okx-chat-bot/provider.js](workers/okx-chat-bot/provider.js)), because a key the provider refuses is indistinguishable from a working one until a buyer's message goes unanswered. The AI credential is an ordered chain rather than one lane (`vertex` on GCP credits, then an Anthropic key, then an Anthropic-wire-format gateway, then an interactive grant, then an OpenAI key): [electProvider()](workers/okx-chat-bot/provider.js) probes every rung on its own env overlay every 15 minutes and serves on the first that answers, so a lane the owner funds is picked up with no deploy and a refused one is never elected. Live state as of 2026-09-09, readable from the `bot_heartbeat` row when `gcloud` is not authenticated: the host has been up since 2026-09-04 with the wallet session logged in and one XMTP client serving, so buyer chat IS delivered, and every lane is refused for lack of funds (Vertex `Lightning dunning decision is deny`, OpenAI `billing_not_active`, OpenRouter `402`), so no reply can be authored. The running revision predates the chain, which its heartbeat meta shows by carrying no `providerLane` / `providerChain` keys. Deploy: [workers/okx-chat-bot/cloudbuild.yaml](workers/okx-chat-bot/cloudbuild.yaml) (`--max-instances=1` is load-bearing, one GCS state writer). Tests: [tests/okx-chat-bot.test.js](tests/okx-chat-bot.test.js) |
| Crypto Data API docs / landing | [pages/crypto.html](pages/crypto.html) `→ /crypto` | In-repo · docs surface | Screenshot-worthy public docs + funnel-top for the free, keyless Crypto Data API (`/api/crypto/*`: token snapshot, security, holders, launches, bonding, whales, symbol, wallet, trending, portfolio, airdrops). Renders hero, endpoint table, tabbed quickstart (curl/JS/Python), per-endpoint request/response reference and a live "Try it" console. Probes production at runtime and marks each endpoint Live vs Coming soon (degrades gracefully while sibling endpoints roll out). CTAs funnel to paid uniques: [/launcher](pages/launcher.html), [/vanity-wallet](public/vanity-wallet.html), [/forge](pages/forge.html). Spec: x402-overhaul campaign (retired 2026-07-28; readable in git history) |
| Free Crypto Data API + discovery | [api/crypto/index.js](api/crypto/index.js) `→ /api/crypto`, [api/crypto/openapi.js](api/crypto/openapi.js) `→ /api/crypto/openapi.json`, docs [docs/crypto-api.md](docs/crypto-api.md) `→ /docs/crypto-api` | Live · docs surface | Front door + OpenAPI 3.1 for the free, keyless Crypto Data API. Catalog assembler [api/_lib/crypto-catalog/index.js](api/_lib/crypto-catalog/index.js) merges `api/_lib/crypto-catalog/*.js` self-describing descriptors (currently `airdrops`, `bonding`, `holders`, `launches`, `portfolio`, `security`, `symbol`, `token`, `trending`, `wallet`, `whales`) via a static import barrel (the production source of truth, since Vercel's bundler only ships what static imports reach) plus an additive dev-time directory glob; skips malformed/throwing entries, never throws, valid-empty when zero. New descriptor = drop the file AND add its import line to the barrel. OpenAPI built from the same catalog via [api/_lib/crypto-catalog/openapi.js](api/_lib/crypto-catalog/openapi.js). Index does HTML/JSON content negotiation. Endpoint handlers live in `api/crypto/*.js`. Spec: x402-overhaul campaign (retired 2026-07-28; readable in git history) |
| Unified Crypto API aggregator — storefront + OpenAPI | Registry [api/v1/_providers.js](api/v1/_providers.js) (`providerCatalog()`), front door [api/v1/x/[...slug].js](api/v1/x/%5B...slug%5D.js) `→ /api/v1/x/<provider>/<endpoint>`, discovery `→ GET /api/v1/x`, storefront [pages/crypto-api.html](pages/crypto-api.html) `→ /crypto-api`, OpenAPI 3.1 [api/openapi-json.js](api/openapi-json.js) `→ /openapi.json`, docs § in [docs/api-reference.md](docs/api-reference.md#unified-api--apiv1x-aggregator), dir guide [api/v1/README.md](api/v1/README.md) | Live | Third-party crypto/DeFi/on-chain APIs (CoinGecko, DefiLlama × 2, Jupiter, DexScreener, Solana RPC, OpenAI-compatible) re-offered as one bill: free per-IP tier → BYOK pass-through → three.ws plan → x402 pay-per-call, in that order. Everything (OpenAPI paths, the `/crypto-api` live table, the docs quota table) renders from `providerCatalog()` at request/page-load time — no hand-enumeration, so a new provider descriptor shows up everywhere with zero page edits. Test: [tests/openapi-aggregator.test.js](tests/openapi-aggregator.test.js). Not to be confused with the separate `/api/crypto/*` bundle below (different campaign, different registry). Spec: x402-catalog campaign (retired 2026-07-28; readable in git history) |
@@ -167,7 +167,7 @@ bottom.
| ASL recognition (webcam signing → text) | worker [workers/model-asl-recognition/](workers/model-asl-recognition), API [api/asl-recognition.js](api/asl-recognition.js) (`GCP_ASL_RECOGNITION_URL`) | Live | The reverse direction of [sign speech](docs/sign-language.md): the browser extracts MediaPipe landmarks from the webcam (video never leaves the device) and the worker transcribes continuous ASL fingerspelling to text on CPU via LiteRT, with a decode-confidence score and an LLM cleanup pass on the API. Model: Kaggle 2023 ASLFR 1st-place (Apache-2.0 code + weights, FSboard CC BY 4.0 corpus, the license-clean chain most ASL models lack). Cloud Run gen2 + `patchelf --clear-execstack` on the LiteRT .so. Next: word-level signs (PopSign retrain). Tests: [test_decode.py](workers/model-asl-recognition/test_decode.py) (schema + decode, no weights) and [test_model_smoke.py](workers/model-asl-recognition/test_model_smoke.py) (real weights end to end, run inside the built image). |
| Smart home: the connect flow and household management | Page [pages/smart-home.html](pages/smart-home.html) `-> /smart-home` + controller [src/home/connect.js](src/home/connect.js), manage view [src/home/manage.js](src/home/manage.js), styles [src/home/home.css](src/home/home.css). Consumes [api/home/index.js](api/home/index.js) (`GET`/`POST /api/home`), [api/home/[id].js](api/home/[id].js), `/api/home/:id/grants` and `/api/home/:id/log`. Reachability rules shared with the server through the [@three-ws/home-bridge](packages/home-bridge) `./url` subpath export | Live | The front door of the Home lane: how a stranger gets from "I have a Home Assistant somewhere" to "my agent is connected", and every way that goes wrong. Fifteen designed states rather than a happy path plus an error toast: signed out, empty, inline validation, the private-host refusal, verifying, connected, a rejected token, unreachable, degraded, disconnected, many homes, the plan ceiling, the dial-out pairing flow, one deep-linked home, and an id that is not this account's. Every one of them names itself on `#hm-root[data-state]`, so "which state am I in" is answerable from outside the page and no state can be reached only by accident. Two rules hold the surface together. **The token goes to the server once and never comes back**: the field is `type=password` with a reveal toggle, it is never written to storage, never put in a URL, never logged and never echoed by the API, and it is gone from the DOM the moment a connect succeeds. **Reachability is decided in the browser, before the network**: a `192.168.x.x` or `.local` address is refused with zero network calls using the same `normalizeBaseUrl`/`isPrivateHost` the server validates with, so the two can never disagree and nobody waits out a timeout to be told something we knew at the first keystroke. Every string that came from a house (a label, an entity id, an area name) is rendered as `textContent`, because those are attacker-influenceable and there is a physical actuator on the other end. Docs: [docs/smart-home.md](docs/smart-home.md). Tests: [tests/e2e/home-connect.spec.js](tests/e2e/home-connect.spec.js) (behaviour on every screen), [tests/e2e/home-connect-gallery.spec.js](tests/e2e/home-connect-gallery.spec.js) (all fifteen photographed at 1440px and 320px, each asserted to log nothing and to fit a 320px screen), [tests/e2e/home-connect-live.spec.js](tests/e2e/home-connect-live.spec.js) (nothing stubbed: a real Home Assistant connected, the real token proven absent from every browser store, and the container really stopped to reach the degraded state) |
| Smart home: plans, entitlements and quotas (`/smart-home/plan`) | Page [pages/smart-home-plan.html](pages/smart-home-plan.html) `-> /smart-home/plan` + controller [src/home/plan.js](src/home/plan.js), styles [src/home/plan.css](src/home/plan.css). Resolver [api/_lib/home/entitlements.js](api/_lib/home/entitlements.js), counters [api/_lib/home/usage.js](api/_lib/home/usage.js), endpoint [api/home/plan.js](api/home/plan.js). Enforced at [api/home/index.js](api/home/index.js) (acquisition), [api/home/[id]/stream.js](api/home/[id]/stream.js) (streams), [api/home/[id]/members.js](api/home/[id]/members.js) (seats), [api/home/[id]/call.js](api/home/[id]/call.js) (paused homes), [api/_lib/home/privacy.js](api/_lib/home/privacy.js) (retention ceiling), [api/_lib/home/turn-gate.js](api/_lib/home/turn-gate.js) called from [api/chat.js](api/chat.js) (agent turns). Cost measurement [scripts/measure-home-entitlement-cost.mjs](scripts/measure-home-entitlement-cost.mjs). Migration `api/_lib/migrations/20260903160000_home_plan_overrides.sql` | Live | What a plan covers for connected houses, and the two things no plan may ever refuse. **A limit never blocks a safety action:** over quota, downgraded or paused, locking a door, closing a garage or valve and arming an alarm always execute, checked before any counter is read and classified from the domain and service alone so the exemption never depends on a healthy socket. **Nothing about the gate is sold:** confirmations, the action log's integrity and every role safety property are on every tier and there is deliberately no dimension for any of them. Seven metered dimensions (homes, seats, streams, voice minutes, agent turns, log retention, relay connections), each carrying the measured cost that justifies it. **No fourth entitlement system:** limits resolve from `account-tier.js` badges merged by MAX (so holding $THREE never costs a Pro their Pro limits) and are scaled by the `three-tier.js` holder multiplier, never by thresholds of our own. **One counter:** an agent turn is the same priced `usage_events` chat row an invoice reads, tagged with `home_id`, not a second tally that could disagree. That row is also what the turn gate reads: over quota the agent refuses a home action, but read-only tools stay open (the model has to see which door is open before it can lock it) and safety actions are never reached by the gate at all. The gate **fails open**, so a billing read that hiccups serves the user rather than locking them out of their own house. **A downgrade pauses, it never disconnects:** excess homes get `deactivated_at` with a written reason, the credential and the action log survive, `revoked_at` stays null, and the user swaps which houses are live. Enterprise limits are a `home_plan_overrides` row with the reason on it, not a tier. Every number is env-overridable (`HOME_LIMIT_<TIER>_<DIMENSION>`), so applying an owner-approved price is a config change. Docs: [docs/home-plans.md](docs/home-plans.md). Tests: [tests/home-entitlements.test.js](tests/home-entitlements.test.js), [tests/home-turn-gate.test.js](tests/home-turn-gate.test.js), [tests/e2e/home-plan.spec.js](tests/e2e/home-plan.spec.js) |
-| Smart home: the live 3D home (`/smart-home/:id`) | Page [pages/home-scene.html](pages/home-scene.html) `-> /smart-home/:id` and `-> /home/:id`, styles [public/home-scene.css](public/home-scene.css). Pure layout [src/home/scene-model.js](src/home/scene-model.js), WebGL renderer [src/home/scene-render.js](src/home/scene-render.js), the 2D house [src/home/scene-fallback.js](src/home/scene-fallback.js), page controller [src/home/scene.js](src/home/scene.js), the top-down floorplan editor the scene follows [src/home/floorplan.js](src/home/floorplan.js) with its validation and storage [api/_lib/home/layout.js](api/_lib/home/layout.js) behind `GET`/`PUT`/`DELETE /api/home/:id/layout` [api/home/[id]/layout.js](api/home/%5Bid%5D/layout.js), browser half of the API contract [src/home/api.js](src/home/api.js). Consumes `GET /api/home/:id`, the SSE stream [api/home/[id]/stream.js](api/home/%5Bid%5D/stream.js), [api/home/[id]/call.js](api/home/%5Bid%5D/call.js) and `/api/home/:id/grants` | Live | A live 3D model of the user's house, driven entirely by their own Home Assistant, with the agent's body standing in it. **Home Assistant stores no geometry**, so everything spatial is derived from its four flat registries by a PURE module with no Three.js in it: floors stack by level, rooms pack per floor in name order, and a saved floorplan overrides the packing through one `layout` option. That split is why the layout of a real recorded house is asserted in a unit test without a GPU. **A room is lit the way the room is lit**: colour and intensity from the lights that are actually on, genuinely dark when they are all off, and the outline around its footprint goes amber the moment anything in it is unlocked or open. **A room with sixty entities is not sixty objects**: objects are ranked (lights, locks, covers, climate, alarms, media, fans, cameras, the rest), capped per surface, and the remainder summarized; numeric sensors become readouts, never objects. A device Home Assistant cannot reach is drawn as a ghost with a marker, never omitted. **The house never empties**: a dropped connection greys the scene, keeps every room, and says how old it is. Stale (the platform is retrying) and disconnected (nobody is) are different pills, and only the second offers a button. **The 2D view is not a consolation prize**: same model, real rows and real buttons, reads and operates the whole house, and it is where a browser with no WebGL or a device measured below 18 fps is routed automatically, with the page saying so. A guarded action answers 409 and the question is rendered on the object it would move, with an optional standing allowance for that one entity. Measured: 86 ms median from a real Home Assistant service call to the SSE frame, 10 to 23 ms from that frame to the painted one, 0.1 to 0.3 ms of per-frame work, 108 draw calls, and a flat heap across ten minutes of live updates. Docs: [docs/home-scene.md](docs/home-scene.md). Tests: [tests/home-scene-model.test.js](tests/home-scene-model.test.js), [tests/e2e/home-scene.spec.js](tests/e2e/home-scene.spec.js) |
+| Smart home: the live 3D home (`/smart-home/:id`) | Page [pages/home-scene.html](pages/home-scene.html) `-> /smart-home/:id` and `-> /home/:id`, styles [public/home-scene.css](public/home-scene.css). Pure layout [src/home/scene-model.js](src/home/scene-model.js), WebGL renderer [src/home/scene-render.js](src/home/scene-render.js), the 2D house [src/home/scene-fallback.js](src/home/scene-fallback.js), page controller [src/home/scene.js](src/home/scene.js), the top-down floorplan editor the scene follows [src/home/floorplan.js](src/home/floorplan.js) with its validation and storage [api/_lib/home/layout.js](api/_lib/home/layout.js) behind `GET`/`PUT`/`DELETE /api/home/:id/layout` [api/home/[id]/layout.js](api/home/%5Bid%5D/layout.js), browser half of the API contract [src/home/api.js](src/home/api.js). Consumes `GET /api/home/:id`, the SSE stream [api/home/[id]/stream.js](api/home/%5Bid%5D/stream.js), [api/home/[id]/call.js](api/home/%5Bid%5D/call.js) and `/api/home/:id/grants` | Live | A live 3D model of the user's house, driven entirely by their own Home Assistant, with the agent's body standing in it. **The body is the visitor's own agent**, resolved through the canonical `my agent` record ([src/agents/active-agent.js](src/agents/active-agent.js)) that the walk world, `/play` and the voice satellite read, swapped live when they switch agents in another tab and falling back to the platform body when they have no agent, no avatar, or a private one (which publishes no model URL by design). **Home Assistant stores no geometry**, so everything spatial is derived from its four flat registries by a PURE module with no Three.js in it: floors stack by level, rooms pack per floor in name order, and a saved floorplan overrides the packing through one `layout` option. That split is why the layout of a real recorded house is asserted in a unit test without a GPU. **A room is lit the way the room is lit**: colour and intensity from the lights that are actually on, genuinely dark when they are all off, and the outline around its footprint goes amber the moment anything in it is unlocked or open. **A room with sixty entities is not sixty objects**: objects are ranked (lights, locks, covers, climate, alarms, media, fans, cameras, the rest), capped per surface, and the remainder summarized; numeric sensors become readouts, never objects. A device Home Assistant cannot reach is drawn as a ghost with a marker, never omitted. **The house never empties**: a dropped connection greys the scene, keeps every room, and says how old it is. Stale (the platform is retrying) and disconnected (nobody is) are different pills, and only the second offers a button. **The 2D view is not a consolation prize**: same model, real rows and real buttons, reads and operates the whole house, and it is where a browser with no WebGL or a device measured below 18 fps is routed automatically, with the page saying so. A guarded action answers 409 and the question is rendered on the object it would move, with an optional standing allowance for that one entity. Measured: 86 ms median from a real Home Assistant service call to the SSE frame, 10 to 23 ms from that frame to the painted one, 0.1 to 0.3 ms of per-frame work, 108 draw calls, and a flat heap across ten minutes of live updates. Docs: [docs/home-scene.md](docs/home-scene.md). Tests: [tests/home-scene-model.test.js](tests/home-scene-model.test.js), [tests/e2e/home-scene.spec.js](tests/e2e/home-scene.spec.js) |
| three.ws Home privacy, retention, export and deletion | Inventory + purge + export + deletion [api/_lib/home/privacy.js](api/_lib/home/privacy.js), disclosure copy [api/_lib/home/disclosure.js](api/_lib/home/disclosure.js), endpoint [api/home/privacy.js](api/home/privacy.js) `-> /api/home/privacy`, schema [api/_lib/migrations/20260903180000_home_privacy_retention.sql](api/_lib/migrations/20260903180000_home_privacy_retention.sql), purge section E of [api/cron/db-retention.js](api/cron/db-retention.js) | Live | The single place that knows everything the Home lane stores. `INVENTORY` is not documentation of the code, it IS the code: the endpoint renders it, the export walks it, and [tests/home-privacy.test.js](tests/home-privacy.test.js) re-derives the set of `home_*` tables from the migration files and fails the build if one of them has no row, which is what stops an undocumented data class from ever landing. The load-bearing promise is that entity names and states are never persisted (the room graph lives in the bridge runtime's memory and dies with the instance); a second test fails any migration that creates a table shaped like a state history. The action log's window is per home and the owner's, 90 days by default, adjustable down to a day and up to ten years with a written reason the schema also enforces, applied immediately on change and swept by the platform's existing retention cron rather than a cron of its own. Deletion goes further than the cascades: memberships and grants on other people's homes, the departed user's id inside another household's action log, invitations addressed to their email, and the lane's rows in the 365-day `audit_log`. Docs: [docs/home-privacy.md](docs/home-privacy.md). Tests: [tests/home-privacy.test.js](tests/home-privacy.test.js) |
| Avatar thumbnail coverage | [api/_lib/avatar-thumbs.js](api/_lib/avatar-thumbs.js) + crons [api/cron/avatar-thumbnail-backfill.js](api/cron/avatar-thumbnail-backfill.js) (absent thumbnails, `*/5`) and [api/cron/avatar-thumbnail-render.js](api/cron/avatar-thumbnail-render.js) (stale marketplace thumbnails, `*/10`) | Live | The single owner of `avatars.thumbnail_key`. Enforces the invariant that a thumbnail key is persisted **only after** its R2 object is confirmed to exist — a key pointing at a missing object makes R2 return 404 `text/plain`, which Chrome blocks as `ERR_BLOCKED_BY_ORB` inside an `<img>`. Fills coverage cheapest-first: zero-copy adoption of `forge_creations.preview_key`, then a headless-chromium GLB→PNG render to `thumb/<avatarId>.png` ([api/_lib/render-glb.js](api/_lib/render-glb.js)). Bounded retries via the `avatar_thumbnail_backfill` claim ledger ([migration](api/_lib/migrations/20260709120000_avatar_thumbnail_backfill.sql)), with `isBrowserInfrastructureError()` separating an OOM-killed browser from a genuinely broken model so a crash never retires healthy avatars. Bulk runner: [scripts/backfill-avatar-thumbnails.mjs](scripts/backfill-avatar-thumbnails.mjs) (`--status`, `--adopt-only`, `--loop`, `--reset-infra`). Docs: [docs/avatar-thumbnails.md](docs/avatar-thumbnails.md). Tests: [tests/avatar-thumbs.test.js](tests/avatar-thumbs.test.js), [tests/home-thumbnail-orb.test.js](tests/home-thumbnail-orb.test.js) |
| Forge model categorization | Classifier [api/_lib/forge-classify.js](api/_lib/forge-classify.js), wired at generation in [api/_lib/forge-store.js](api/_lib/forge-store.js) (`createCreation`), backfill [scripts/backfill-forge-categories.mjs](scripts/backfill-forge-categories.mjs), coloured badges in [src/forge-showcase.js](src/forge-showcase.js) + `/forge-studio` twin | Live | Assigns every forge model a `model_category` (avatar/accessory/item/scene/creature/vehicle/other) from its prompt with a deterministic, dependency-free keyword classifier — most-specific-first so "a knight riding a horse" is an avatar, not a creature. Nothing set the category before, so ~99.8% of the corpus was `other` and the category badges, filters, and Forge-Off category dimension were dead; the classifier now runs inline on every generation (all lanes) and reclassifies ~85% of the 9.8k-row back catalogue (dry-run verified). Coloured category pills render across the showcase/gallery/community feed. Tests: [tests/forge-classify.test.js](tests/forge-classify.test.js). |
@@ -284,6 +284,7 @@ bottom.
| Alpha-drip (a leader prices their signal latency) | Ladder rules [api/\_lib/alpha-drip.js](api/_lib/alpha-drip.js), fanout half [api/\_lib/copy-drip-runtime.js](api/_lib/copy-drip-runtime.js), measured edge half-life [api/\_lib/alpha-drip-stats.js](api/_lib/alpha-drip-stats.js), endpoint [api/copy/alpha-drip.js](api/copy/alpha-drip.js) `-> /api/copy/alpha-drip`, leader editor [src/alpha-drip-panel.js](src/alpha-drip-panel.js) mounted on [/dashboard/copy](src/dashboard-next/pages/copy.js), copier seat on the trader page [src/copy-panel.js](src/copy-panel.js), schema [api/\_lib/migrations/20260903235500\_copy\_alpha\_drip.sql](api/_lib/migrations/20260903235500_copy_alpha_drip.sql) | Live (off unless a leader turns it on) | The $THREE sink on the copy rail: a leader sets how long each holder tier waits before a copy intent is revealed to them, so the thing being sold is latency rather than only a share of the profit. Two rules are structural and enforced server-side: a higher tier can never wait longer than a lower one, and **a drip delays the reveal, never the record** (the intent row is written in full at fanout time and the public track record is untouched, so there are no hidden trades). The copier is shown the seat their balance buys BEFORE they subscribe, not after; a locked intent renders as a countdown that unlocks in place, and `POST /api/copy/executions` refuses an unreleased one in the same statement that changes its status, so the guard cannot be raced. Tier size caps are applied after the copy engine has sized and gated the order, never before, so the copier's own caps still bind first. The edge half-life the fairness warning judges a ladder against is measured from the leader's own profitable closes, and is null rather than guessed below five of them. Release runs inside [api/cron/copy-fanout.js](api/cron/copy-fanout.js) rather than a second cron, so the release cadence is the fanout cadence. Doc: [docs/alpha-drip.md](docs/alpha-drip.md). Tests: [tests/alpha-drip.test.js](tests/alpha-drip.test.js) |
| Ghost-copy (paper-copy a leader) | [pages/ghost-copy.html](pages/ghost-copy.html) + [src/ghost-copy.js](src/ghost-copy.js) `→ /ghost-copy`, simulator [api/_lib/ghost-copy.js](api/_lib/ghost-copy.js), endpoint [api/pump/ghost-copy.js](api/pump/ghost-copy.js), tests [tests/ghost-copy.test.js](tests/ghost-copy.test.js) | Live (paper only, zero custody) | The bridge between the leaderboard and actually copying someone. Pick a verified trader agent, a budget and a window, and it replays that agent's real closed on-chain round-trips (`agent_sniper_positions`) against a wallet that does not exist, sizing every order through `planCopyOrder`, the SAME function the live [copy-fanout cron](api/cron/copy-fanout.js) uses, so the result is what the copy engine would actually have generated rather than a marketing figure. Deliberately unflattering: P&L per trade is derived from exact chain lamports (never the stored pct column), losing round-trips all count, the ghost wallet has a real cash constraint so capital locked in an open copy blocks the next one, every skip carries a machine reason plus a human sentence, and the leader's still-open positions are marked at THEIR last on-chain quote and reported as unrealized outside the realized headline. Needs no account, no wallet, no signature and no funded signer, so it ships ahead of any custody decision; state lives in the query string so a result is deep-linkable and shareable. Converts into [/vaults](pages/vaults.html) for real money. Doc: [docs/ghost-copy.md](docs/ghost-copy.md). |
| Trader Wrapped (season recap) | [pages/wrapped.html](pages/wrapped.html) + [src/wrapped.js](src/wrapped.js) `→ /wrapped`, deck builder [api/_lib/wrapped.js](api/_lib/wrapped.js), endpoint [api/pump/wrapped.js](api/pump/wrapped.js), social card [api/wrapped-og.js](api/wrapped-og.js), share page [api/wrapped-share.js](api/wrapped-share.js) `→ /wrapped/<agent_id>/share`, tests [tests/wrapped.test.js](tests/wrapped.test.js) | Live (read-only, zero custody) | The artifact the trading stack was missing: one image and one link that say what a season was. Eight slides cut server-side from a trader agent's closed on-chain round-trips (scoreboard, best trade, worst trade, the coin that carried it, rhythm, rank against the field with the nearest rival, receipt), swipeable on touch and keyboard-driven on desktop, with the current slide in the URL hash so a share link lands on the slide the sharer was reading. Arithmetic is delegated wholesale to `computeTraderMetrics` in [api/_lib/trader-stats.js](api/_lib/trader-stats.js), the same truth layer the leaderboard and trader profile use, so a recap can never disagree with the profile it links to. Deliberately unflattering: the worst-trade slide is not optional, the scoreboard carries an explicit verdict and colors a red season red, per-coin ranking includes coins that lost money, self-launched coins stay excluded from every superlative, the rank always prints the size of the real field beside it, and under 3 settled round-trips the page says there is no season rather than drawing a chart out of noise. Needs no account, wallet, signature or funded signer. Converts through [Fork](src/fork-trade.js) on the best-trade slide and links onward to the verified profile, [Ghost-copy](pages/ghost-copy.html) and the rival's own wrapped. Doc: [docs/trader-wrapped.md](docs/trader-wrapped.md). |
+| Rivalries (what changed on the trader board) | Engine [api/_lib/rivalries.js](api/_lib/rivalries.js), endpoint [api/sniper/rivalries.js](api/sniper/rivalries.js) `→ /api/sniper/rivalries`, grudge-match strip [src/play/rivalries.js](src/play/rivalries.js) mounted under the leaderboard on [pages/play/arena.html](pages/play/arena.html) `→ /play/arena`, tests [tests/rivalries.test.js](tests/rivalries.test.js) | Live (read-only) | The layer a leaderboard never has: not who is ahead, but who just took whose place. Every trader on the board is paired with the one directly above them and the pair is labelled `overtake`, `debut` or `chase`, with the P&L both sides booked inside the lookback as the reason. Derived from ONE position fetch through the shared `computeTraderMetrics`, run three times: the current window, the same window measured at the lookback cutoff, and the round-trips closed inside the lookback. The past board is re-derived from close timestamps rather than read from a snapshot table, so there is no new table and no cron, it works retroactively from the day it shipped, and it cannot drift out of agreement with `/api/sniper/leaderboard` or with either profile it links to. The copy is assembled from those numbers rather than generated, which keeps a public strip instant, free and checkable: it will not call a trader "behind" when the board ranks on score and that trader actually holds more SOL, and it will not say someone "booked 0.000 SOL" when they simply were not in the market. Ranking boards apply a three settled round-trip floor; the momentum board does not, because one settled trade is the movement it exists to show. Doc: [docs/rivalries.md](docs/rivalries.md). |
| Clip Director (content engine) | [pages/clip-director.html](pages/clip-director.html) + [src/clip-director-page.js](src/clip-director-page.js) `→ /clip-director`, brain [api/_lib/clip-director.js](api/_lib/clip-director.js), endpoint [api/clip-director.js](api/clip-director.js) | Live | Turns one real closed trade (`agent_sniper_positions`) into the optimal shareable card per surface (X / Telegram / Feed): honest hook, the real number, an avatar gesture mapped to a real animation clip, CTA, and alt text, plus an on-chain proof link. Losses get an honest card too. Free-first `llmComplete` with a deterministic same-shape fallback. Doc: [docs/clip-director.md](docs/clip-director.md). |
| Atlas (site-wide Cmd+K search + route map) | Palette [public/atlas.js](public/atlas.js) `-> /atlas.js`, ranking [public/atlas/score.js](public/atlas/score.js), map page [pages/atlas.html](pages/atlas.html) + [src/atlas-page.js](src/atlas-page.js) `-> /atlas`, index [scripts/build-atlas-index.mjs](scripts/build-atlas-index.mjs) `-> public/atlas-index.json`, task catalog [data/atlas-intents.json](data/atlas-intents.json), coverage [scripts/inject-atlas.mjs](scripts/inject-atlas.mjs) | Live | **Discovery layer for 650+ routes.** Cmd+K (or `/`) opens a search over every page in `data/pages.json` plus a curated set of TASKS that answer with ordered steps rather than one link, because "which page starts the get-paid flow" is a harder question than "where is /x402/studio". Zero-dependency single file so it runs identically on 250+ standalone HTML entry points; the index is fetched on first open, not page load. Every task step target is resolved against the real route table at build time, so an onboarding path cannot dead-end on a 404. Injected by a Vite `three-ws-atlas` plugin plus a post-build `dist/` walk (same belt-and-suspenders as [scripts/inject-tour-boot.mjs](scripts/inject-tour-boot.mjs), for the same plugin-ordering reason). Embeds and the nav/footer fragments are excluded. Doc: [docs/atlas.md](docs/atlas.md). Tests: [tests/atlas.test.js](tests/atlas.test.js). |
| Internationalization (i18n) | Runtime [src/i18n.js](src/i18n.js) `→ /i18n.js`; catalogs [public/locales/](public/locales) (`en` + 84 locales, `manifest.json`, `localized-pages.json`); pipeline [scripts/i18n-annotate.mjs](scripts/i18n-annotate.mjs) · [i18n-extract.mjs](scripts/i18n-extract.mjs) · [i18n-translate.mjs](scripts/i18n-translate.mjs) (Vertex Gemini, glossary-locked) · [i18n-backfill.mjs](scripts/i18n-backfill.mjs) (re-translates copy that was baked to English during a backend outage); hreflang in [api/sitemap/[type].js](api/sitemap/%5Btype%5D.js) | Live (85 languages) | LobeHub-model client-side copy swap. `npm run i18n:annotate` auto-injects `data-i18n*` into static HTML (byte-surgical), `i18n:extract` builds `en.json`, `i18n:translate` fans it out to every locale via Vertex with brand/protocol terms masked. `i18n:backfill` finds keys whose value is still byte-identical to English (the silent failure mode: a baked fallback is present, so every later run skips it) and re-translates them. Runtime auto-detects locale, mounts a floating `<lang-switcher>`, and injects hreflang. Doc: [docs/i18n.md](docs/i18n.md). Tests: [tests/i18n.test.js](tests/i18n.test.js), [tests/i18n-annotate.test.js](tests/i18n-annotate.test.js), [tests/i18n-backfill.test.js](tests/i18n-backfill.test.js). |
diff --git a/data/pages.json b/data/pages.json
index af4e54181..79403ec35 100644
--- a/data/pages.json
+++ b/data/pages.json
@@ -3449,6 +3449,14 @@
"changefreq": "monthly",
"added": "2026-07-18"
},
+ {
+ "path": "/docs/rivalries",
+ "title": "Docs · Rivalries",
+ "description": "What changed on the trader board: who passed whom, who just arrived, and who is closing. Derived from close timestamps rather than a snapshot table, so the matchups work retroactively and cannot disagree with the leaderboard.",
+ "priority": 0.6,
+ "changefreq": "monthly",
+ "added": "2026-09-09"
+ },
{
"path": "/docs/clash",
"title": "Docs · Coin Clash",
diff --git a/docs/start-here.md b/docs/start-here.md
index 56b981903..db67c0be5 100644
--- a/docs/start-here.md
+++ b/docs/start-here.md
@@ -144,6 +144,7 @@ Want the model in your hands rather than on a screen? [Materialize](./materializ
- **Following a trader instead of picking?** → [Copy trading](./copy-trading.md): follow a leader two ways, either your agent mirroring real Solana trades inside its own spend policy, or sized intents you sign yourself. Nothing is simulated, so the doc covers the leaderboard weighting, the Smart Money directory, and how performance fees accrue
- **Found a coin and want to act on it now?** → [Fork a trade](./fork-trade.md): one tap on any coin across the Radar, the Live Trade Feed, the Smart Money Radar, or a ghost-copy replay opens the real pump.fun trade panel for it, pre-filled with the size being forked and signed by your own wallet. No custody, no delegated key, the same safety firewall as every other trade surface, and a `?fork=` link so a coin can travel to X or Telegram and open the panel on arrival
- **Not ready to risk money on a trader yet?** → [Ghost-copy](./ghost-copy.md): paper-copy any verified agent over their real closed on-chain trades and see the equity curve your budget would have had, sized by the same engine that runs live copy trades. No wallet, no signature, no account, and every trade your budget could not reach is listed with the reason
+- **Want to know what just changed on the board?** → [Rivalries](./rivalries.md): every trader on the arena leaderboard is paired with the one directly above them and labelled an overtake, a debut or a standing chase, with the profit both sides settled in the last 24 hours as the reason. Worked out by rebuilding what the board said at the cutoff from the trade timestamps, so it reaches back over history already recorded and never disagrees with the leaderboard
- **Watching identities land on-chain?** → [On-chain deployments](./deployments.md): the live feed of agent identities registering on-chain as it happens, Solana Metaplex Core mints alongside ERC-8004 registrations across 15 EVM chains, with chain footprint and capability stats
- **Building against the API?** → [The developer platform](./developer-platform.md): API keys, usage metrics, signed webhooks, the one-call MCP connection test, and the public tool and skills catalogs, with a runnable example per endpoint
- **Auditing an agent's decisions?** → [The Reasoning Ledger](./reasoning-ledger.md): hash-chained, on-chain-anchored records of every consequential agent decision, verifiable by anyone at /ledger
diff --git a/public/changelog-recent.json b/public/changelog-recent.json
index 076268600..08a24ea96 100644
--- a/public/changelog-recent.json
+++ b/public/changelog-recent.json
@@ -1,11 +1,11 @@
{
- "generated_at": "2026-09-09T15:09:28.073Z",
+ "generated_at": "2026-09-09T15:18:36.394Z",
"generated_by": "scripts/build-page-index.mjs from data/pages.json + data/changelog.json",
"site": {
"name": "three.ws",
"url": "https://three.ws"
},
- "total": 3640,
+ "total": 3642,
"complete": false,
"full_feed": "/changelog.json",
"entries": [
@@ -124,6 +124,16 @@
"improvement"
]
},
+ {
+ "date": "2026-09-09",
+ "type": "update",
+ "title": "Fact-check now reads the part of a source that is actually about your claim",
+ "summary": "When the fact-checker fell back to Wikipedia, it only ever read each article's opening summary. An opening summary says what a page is about, so it carries the answer only when your claim is about the page's subject. Ask whether Napoleon was unusually short and it read the first lines of \"Napoleon\", \"Napoleon III\" and \"Napoleonic Wars\", none of which mention height at all, so every source was scored as saying nothing and the verdict came back \"not enough evidence\" while the passage about his height sat further down the same article. The checker now reads each article in full and picks the paragraphs that actually discuss your claim, so that question returns the \"Appearance and image\" and \"Napoleon's height\" sections instead. Reference lists are skipped, since they repeat your words without asserting anything, and the deeper read is bounded by the time the check already had, so a slow source can never cost you the evidence that already arrived.",
+ "link": "/docs/fact-check",
+ "tags": [
+ "improvement"
+ ]
+ },
{
"date": "2026-09-09",
"type": "update",
@@ -145,6 +155,16 @@
"infra"
]
},
+ {
+ "date": "2026-09-09",
+ "type": "update",
+ "title": "News cards stop losing their preview picture to a cold cache",
+ "summary": "About a fifth of the crypto-news feed ships text-only, so those cards start as a source tile and upgrade in place once the publisher's own preview image is found. That lookup searched all 197 publisher feeds at once and gave up after 2.5 seconds, which on a freshly started server dropped whichever feed had not answered yet, so an article the feed had just served came back as unknown. The reader kept its tile, but the failed lookup was recorded as a public answer for the next ten minutes, so every visitor to /markets and /markets/news got the same dead result and a 404 in their browser console. The lookup now asks the one publisher that could hold the article first and waits for it, and a result that could not be confirmed is held for 30 seconds instead of ten minutes.",
+ "link": "/markets/news",
+ "tags": [
+ "fix"
+ ]
+ },
{
"date": "2026-09-09",
"type": "update",
@@ -1591,27 +1611,6 @@
"feature",
"security"
]
- },
- {
- "date": "2026-09-04",
- "type": "update",
- "title": "Agent wallets can pay three.ws on X Layer again",
- "summary": "Buying agents that pay for a 3D generation on X Layer sign their payment with a wallet that carries an on-chain delegation, which is how every OKX agent wallet is built. Our checker judged those signatures with a contract call that answers a different question, so it turned away correct, fully funded payments and quoted the price a second time instead of doing the work. Payments are now checked the same way the stablecoin itself checks them, so a signed authorization that the network would honour is one we accept.",
- "link": null,
- "tags": [
- "fix"
- ]
- },
- {
- "date": "2026-09-04",
- "type": "update",
- "title": "Cross-origin 3D loads no longer break on a header your site never sent",
- "summary": "Embedding a three.ws avatar on your own site loads the model straight from us, and the site answers every origin. One case still failed: if your loader attached any extra header (a trace id, an auth token, anything a browser does not treat as standard), the browser first asked us for permission and we answered with a 404 page, so the model never loaded. That request is now answered properly, and the answer echoes exactly the headers you asked about. Alongside it, the media docs now spell out the three ways to read a model file and which one to pick: files under three.ws load directly, generated assets load directly from three.ws/cdn, and the /api/glb proxy is there for a URL whose host you cannot identify.",
- "link": "/docs/media-api",
- "tags": [
- "fix",
- "docs"
- ]
}
]
}
diff --git a/public/changelog.json b/public/changelog.json
index d8bb0f7c3..3abf705b7 100644
--- a/public/changelog.json
+++ b/public/changelog.json
@@ -1,5 +1,5 @@
{
- "generated_at": "2026-09-09T15:09:28.022Z",
+ "generated_at": "2026-09-09T15:18:36.344Z",
"generated_by": "scripts/build-page-index.mjs from data/pages.json + data/changelog.json",
"site": {
"name": "three.ws",
@@ -121,6 +121,16 @@
"improvement"
]
},
+ {
+ "date": "2026-09-09",
+ "type": "update",
+ "title": "Fact-check now reads the part of a source that is actually about your claim",
+ "summary": "When the fact-checker fell back to Wikipedia, it only ever read each article's opening summary. An opening summary says what a page is about, so it carries the answer only when your claim is about the page's subject. Ask whether Napoleon was unusually short and it read the first lines of \"Napoleon\", \"Napoleon III\" and \"Napoleonic Wars\", none of which mention height at all, so every source was scored as saying nothing and the verdict came back \"not enough evidence\" while the passage about his height sat further down the same article. The checker now reads each article in full and picks the paragraphs that actually discuss your claim, so that question returns the \"Appearance and image\" and \"Napoleon's height\" sections instead. Reference lists are skipped, since they repeat your words without asserting anything, and the deeper read is bounded by the time the check already had, so a slow source can never cost you the evidence that already arrived.",
+ "link": "/docs/fact-check",
+ "tags": [
+ "improvement"
+ ]
+ },
{
"date": "2026-09-09",
"type": "update",
@@ -142,6 +152,16 @@
"infra"
]
},
+ {
+ "date": "2026-09-09",
+ "type": "update",
+ "title": "News cards stop losing their preview picture to a cold cache",
+ "summary": "About a fifth of the crypto-news feed ships text-only, so those cards start as a source tile and upgrade in place once the publisher's own preview image is found. That lookup searched all 197 publisher feeds at once and gave up after 2.5 seconds, which on a freshly started server dropped whichever feed had not answered yet, so an article the feed had just served came back as unknown. The reader kept its tile, but the failed lookup was recorded as a public answer for the next ten minutes, so every visitor to /markets and /markets/news got the same dead result and a 404 in their browser console. The lookup now asks the one publisher that could hold the article first and waits for it, and a result that could not be confirmed is held for 30 seconds instead of ten minutes.",
+ "link": "/markets/news",
+ "tags": [
+ "fix"
+ ]
+ },
{
"date": "2026-09-09",
"type": "update",
diff --git a/public/changelog.xml b/public/changelog.xml
index 0fd1da07a..19fd0eb56 100644
--- a/public/changelog.xml
+++ b/public/changelog.xml
@@ -94,6 +94,14 @@
<category>improvement</category>
<description>When the fact-checker fell back to Wikipedia, it only ever read each article's opening summary. An opening summary says what a page is about, so it carries the answer only when your claim is about the page's subject. Ask whether Napoleon was unusually short and it read the first lines of &quot;Napoleon&quot;, &quot;Napoleon III&quot; and &quot;Napoleonic Wars&quot;, none of which mention height at all, so every source was scored as saying nothing and the verdict came back &quot;not enough evidence&quot; while the passage about his height sat further down the same article. The checker now reads each article in full and picks the paragraphs that actually discuss your claim, so that question returns the &quot;Appearance and image&quot; and &quot;Napoleon's height&quot; sections instead. Reference lists are skipped, since they repeat your words without asserting anything, and the deeper read is bounded by the time the check already had, so a slow source can never cost you the evidence that already arrived.</description>
</item>
+ <item>
+ <title>Fact-check now reads the part of a source that is actually about your claim</title>
+ <link>https://three.ws/docs/fact-check</link>
+ <guid isPermaLink="false">2026-09-09:Fact-check now reads the part of a source that is actually about your claim</guid>
+ <pubDate>Wed, 09 Sep 2026 12:00:00 GMT</pubDate>
+ <category>improvement</category>
+ <description>When the fact-checker fell back to Wikipedia, it only ever read each article's opening summary. An opening summary says what a page is about, so it carries the answer only when your claim is about the page's subject. Ask whether Napoleon was unusually short and it read the first lines of &quot;Napoleon&quot;, &quot;Napoleon III&quot; and &quot;Napoleonic Wars&quot;, none of which mention height at all, so every source was scored as saying nothing and the verdict came back &quot;not enough evidence&quot; while the passage about his height sat further down the same article. The checker now reads each article in full and picks the paragraphs that actually discuss your claim, so that question returns the &quot;Appearance and image&quot; and &quot;Napoleon's height&quot; sections instead. Reference lists are skipped, since they repeat your words without asserting anything, and the deeper read is bounded by the time the check already had, so a slow source can never cost you the evidence that already arrived.</description>
+ </item>
<item>
<title>Fade Radar: the wallets that have never once been right, and the coins they are crowding into</title>
<link>https://three.ws/fade</link>
@@ -110,6 +118,14 @@
<category>fix</category>
<description>If your Home Assistant is only reachable inside your house, three.ws reaches it through the three.ws integration you install there, which dials out to us. Those homes were building their room list once, when the connection opened, and then quietly ignoring changes: assign a lamp to the kitchen, rename a floor, add a device, and the 3D house kept showing the old arrangement until the connection happened to be rebuilt. Homes connected by a direct address never had this problem, and the split was in the relay, which was not carrying the four small notices Home Assistant sends when its rooms, floors, devices or entities change. It carries them now, so both kinds of connection track the house the same way and updates land in seconds. The notices only say that something changed; the relay already read the full room and floor lists, so nothing new about your house is exposed by this.</description>
</item>
+ <item>
+ <title>News cards stop losing their preview picture to a cold cache</title>
+ <link>https://three.ws/markets/news</link>
+ <guid isPermaLink="false">2026-09-09:News cards stop losing their preview picture to a cold cache</guid>
+ <pubDate>Wed, 09 Sep 2026 12:00:00 GMT</pubDate>
+ <category>fix</category>
+ <description>About a fifth of the crypto-news feed ships text-only, so those cards start as a source tile and upgrade in place once the publisher's own preview image is found. That lookup searched all 197 publisher feeds at once and gave up after 2.5 seconds, which on a freshly started server dropped whichever feed had not answered yet, so an article the feed had just served came back as unknown. The reader kept its tile, but the failed lookup was recorded as a public answer for the next ten minutes, so every visitor to /markets and /markets/news got the same dead result and a 404 in their browser console. The lookup now asks the one publisher that could hold the article first and waits for it, and a result that could not be confirmed is held for 30 seconds instead of ten minutes.</description>
+ </item>
<item>
<title>Oracle: a failed request now says so, instead of claiming no coins have been scored</title>
<link>https://three.ws/oracle</link>
diff --git a/public/features.json b/public/features.json
index bcb047905..ce07536dc 100644
--- a/public/features.json
+++ b/public/features.json
@@ -1,5 +1,5 @@
{
- "generated_at": "2026-09-09T07:20:21.637Z",
+ "generated_at": "2026-09-09T15:18:36.257Z",
"generated_by": "scripts/build-page-index.mjs from data/pages.json",
"site": {
"name": "three.ws",
@@ -4208,6 +4208,16 @@
"tags": [],
"showcase": false
},
+ {
+ "path": "/docs/rivalries",
+ "title": "Docs · Rivalries",
+ "description": "What changed on the trader board: who passed whom, who just arrived, and who is closing. Derived from close timestamps rather than a snapshot table, so the matchups work retroactively and cannot disagree with the leaderboard.",
+ "added": "2026-09-09",
+ "auth": null,
+ "indexable": true,
+ "tags": [],
+ "showcase": false
+ },
{
"path": "/docs/clash",
"title": "Docs · Coin Clash",
diff --git a/public/llms-full.txt b/public/llms-full.txt
index e178db91c..c7181dfa3 100644
--- a/public/llms-full.txt
+++ b/public/llms-full.txt
@@ -2572,6 +2572,12 @@ URL: https://three.ws/docs/radar
Live pump.fun launch intelligence: every new coin scored in its first 90 seconds for bundle vs organic, wallet concentration, dev behaviour, and risk flags.
+### Docs · Rivalries
+
+URL: https://three.ws/docs/rivalries
+
+What changed on the trader board: who passed whom, who just arrived, and who is closing. Derived from close timestamps rather than a snapshot table, so the matchups work retroactively and cannot disagree with the leaderboard.
+
### Docs · Coin Clash
URL: https://three.ws/docs/clash
diff --git a/public/llms.txt b/public/llms.txt
index e0fb05212..4cdd57f6f 100644
--- a/public/llms.txt
+++ b/public/llms.txt
@@ -444,6 +444,7 @@ Docs, tutorials, and reference.
- [Docs · Smart Money Radar](https://three.ws/docs/smart-money): A live first-party reputation graph of every pump.fun wallet: cross each coin's buyers with the coins that actually graduated to surface the wallets that keep picking winners.
- [Docs · Fade Radar](https://three.ws/docs/fade-radar): How the Fade Radar scores a coin's buy side: what makes a wallet a reverse indicator, the 0 to 100 fade score, the verdict bands, and the out-of-sample measurement behind them.
- [Docs · Coin Radar](https://three.ws/docs/radar): Live pump.fun launch intelligence: every new coin scored in its first 90 seconds for bundle vs organic, wallet concentration, dev behaviour, and risk flags.
+- [Docs · Rivalries](https://three.ws/docs/rivalries): What changed on the trader board: who passed whom, who just arrived, and who is closing. Derived from close timestamps rather than a snapshot table, so the matchups work retroactively and cannot disagree with the leaderboard.
- [Docs · Coin Clash](https://three.ws/docs/clash): Token-gated community warfare: every coin community is an army. Hold the coin, enlist, and rally for your faction in timed on-chain battles.
- [Docs · x402 Studio](https://three.ws/docs/x402-studio): The merchant console for running a paid x402 business: products and pricing, payout and agent wallets, real USDC send and receive with .sol resolution, a storefront and pay-button builder, giving, and security.
- [Docs · Fact Check](https://three.ws/docs/fact-check): Sourced fact-checking with cryptographic content attestations, a published accuracy benchmark, and a free daily quota then pay-per-check via x402.
diff --git a/public/sitemap/index.html b/public/sitemap/index.html
index 1acc2e031..28b5c7008 100644
--- a/public/sitemap/index.html
+++ b/public/sitemap/index.html
@@ -96,7 +96,7 @@
<main class="sm-wrap">
<div class="sm-hero">
<h1>Sitemap</h1>
- <p>All 926 pages on three.ws, grouped by purpose. Looking for the machine-readable versions?</p>
+ <p>All 927 pages on three.ws, grouped by purpose. Looking for the machine-readable versions?</p>
<div class="sm-formats">
<a href="/sitemap.xml"><code>sitemap.xml</code> · for search engines</a>
<a href="/llms.txt"><code>llms.txt</code> · for AI agents</a>
@@ -108,7 +108,7 @@
</div>
<div class="sm-filter">
<svg viewBox="0 0 24 24" fill="none" stroke="currentColor" stroke-width="2" stroke-linecap="round" aria-hidden="true"><circle cx="11" cy="11" r="7"/><path d="m21 21-4.3-4.3"/></svg>
- <input id="sm-filter-input" type="search" placeholder="Filter 926 pages by name, path, or description…" aria-label="Filter pages" autocomplete="off" spellcheck="false" />
+ <input id="sm-filter-input" type="search" placeholder="Filter 927 pages by name, path, or description…" aria-label="Filter pages" autocomplete="off" spellcheck="false" />
<kbd id="sm-filter-kbd" aria-hidden="true">/</kbd>
</div>
<p class="sm-filter-count" role="status" aria-live="polite"><span id="sm-filter-count"></span><span id="sm-filter-hint" aria-hidden="true"></span></p>
@@ -132,6 +132,14 @@
<span class="sm-desc">How the Fade Radar scores a coin's buy side: what makes a wallet a reverse indicator, the 0 to 100 fade score, the verdict bands, and the out-of-sample measurement behind them.</span>
</a>
</li>
+ <li>
+ <a href="/docs/rivalries">
+ <span class="sm-new-date">Sep 9, 2026</span>
+ <span class="sm-title">Docs · Rivalries</span>
+ <span class="sm-path">/docs/rivalries</span>
+ <span class="sm-desc">What changed on the trader board: who passed whom, who just arrived, and who is closing. Derived from close timestamps rather than a snapshot table, so the matchups work retroactively and cannot disagree with the leaderboard.</span>
+ </a>
+ </li>
<li>
<a href="/globe">
<span class="sm-new-date">Sep 8, 2026</span>
@@ -180,14 +188,6 @@
<span class="sm-desc">Every asset a Robinhood Chain portfolio may hold, priced live: tokenized equities, crypto majors and chain-native tokens with on-chain liquidity, 24h volume, price change and the NAV premium against each equity's Chainlink feed. Sortable and filterable by asset class.</span>
</a>
</li>
- <li>
- <a href="/docs/seo">
- <span class="sm-new-date">Sep 7, 2026</span>
- <span class="sm-title">Docs · SEO and crawler rendering</span>
- <span class="sm-path">/docs/seo</span>
- <span class="sm-desc">How a JavaScript site is served to search engines: the four server-side render paths, the contract that keeps a page and its sitemap entry agreeing, and how to audit the live origin.</span>
- </a>
- </li>
</ul>
</section>
<nav class="sm-toc" aria-label="Sections">
@@ -2555,7 +2555,7 @@
<section class="sm-section" id="learn">
<header>
- <h2>Learn<span class="sm-count" data-count>426</span></h2>
+ <h2>Learn<span class="sm-count" data-count>427</span></h2>
<p class="sm-section-desc">Docs, tutorials, and reference.</p>
</header>
<ul class="sm-list">
@@ -3168,6 +3168,13 @@
<span class="sm-desc">Live pump.fun launch intelligence: every new coin scored in its first 90 seconds for bundle vs organic, wallet concentration, dev behaviour, and risk flags.</span>
</a>
</li>
+ <li>
+ <a href="/docs/rivalries">
+ <span class="sm-title">Docs · Rivalries</span>
+ <span class="sm-path">/docs/rivalries</span>
+ <span class="sm-desc">What changed on the trader board: who passed whom, who just arrived, and who is closing. Derived from close timestamps rather than a snapshot table, so the matchups work retroactively and cannot disagree with the leaderboard.</span>
+ </a>
+ </li>
<li>
<a href="/docs/clash">
<span class="sm-title">Docs · Coin Clash</span>