Garry's Opinionated OpenClaw/Hermes Agent Brain
TypeScript
Merge remote-tracking branch 'origin/master' into capy/gbra224-w3-self-heals
garrytan3a28d2a1d37d
diff --cc CHANGELOG.mdindex a945bfdd0d,13fc31a1c0..7efb46c4ed--- a/CHANGELOG.md+++ b/CHANGELOG.md@@@ -10,29 -10,18 +10,41 @@@ credits are retained; no result has beeidentifiers and attribution are available in the pre-removal Git revision`6040075c6cb95be5881cc2e1b76ef7d71f4e5d29` (retained on 2026-09-23).+## [Unreleased]++**Dream and embedding maintenance fix their own routine noise for free, and tell you only when that stops working.**++A self-hosted nightly cycle could end `partial` for reasons that needed no one: a purge or consolidate write accepted but still committing after its 5-second wait, a synthesize window with nothing new in it, or pages whose vectors already belonged to the current model but carried no signature, so they read as another model's space and invited a paid re-embed. Each of these now heals inside the cycle with no model or provider call, within a fixed bound, with a record of what it did and a config switch to turn it off. When a heal hits its bound, a doctor check says so.++### What changed++| Situation | Before | Now |+|---|---|---|+| A purge or consolidate write is accepted but still pending after the publish wait | the phase failed and the cycle ended `partial` | the phase warns `dream_write_deferred` with the receipt id; the next pass reads that receipt before it submits anything for the source. After 3 passes still pending, doctor `dream_write_deferred` names the receipt |+| Every transcript in the synthesize window is already synthesized | triage, the manifest context and the child queue still ran | the phase returns `no_candidates` before any of that |+| A transcript grows between runs that keep dying | each run had a new content key, so the 3-deaths breaker never tripped | deaths also count per transcript (`dream:synth-v2:@file:<source>:<file>`); the same 3-in-24h limit applies |+| `gbrain dream reset-key` on a `dream:phase:` key that `--list` printed | refused as not a breaker key | resets it |+| Pages whose chunks already hold current-model vectors but have no or an old signature | read as another model's space until a paid drain stamped them | the embed phase stamps them first, up to 500 pages a run, with no provider call |+| A soft-deleted page keeps unembedded chunks | health's embedding coverage dropped | coverage counts live pages only, like `missing_embeddings` |++### For operators++- New doctor checks `dream_write_deferred` and `embedding_restamp_stalled` (read-only). Both are meant for the page class of `gbrain doctor --alert`.+- New switches, on by default: `dream.write_deferral.enabled` and `embed.restamp.enabled`. Setting either to `false` restores the old behavior.+- Every heal writes a row to the `heal.audit` config record (last 200 rows, 30 days).+- Nothing runs on upgrade: the restamp runs only in the maintenance cycle's embed phase, and repo-less brains get the same behavior.+ ## [0.60.477.0] - 2026-10-28++ **A trust release confirms only what its preview showed.**++ Releasing a held fact that already sits in a page's `## Facts` table confirmed that row's confidence, notability, validity and other cells, though the owner's preview showed only the text (#7656). The database-only release, which every brain without a repository uses, could also release a hold that was re-gated after the owner approved it (#7657).++ ### What changed++ - **The preview names every field the release confirms.** That's the held row's cells as they stand, or the held payload's kind, visibility, validity, source and context. The approval binds them, so a row edited after the approval refuses `preview_changed` and the fact stays held.+ - **The preview can't be made to show a field the row doesn't have.** A held fact's text is printed quoted and escaped, any other value containing a quote, comma or semicolon is quoted the same way, and invisible or bidirectional formatting characters become spaces.+ - **The database-only release re-checks the approval** on the locked hold and refuses `preview_changed` when the hold changed after the owner approved it, the same as the managed path.+ - **New doctor check `trust_release_unpreviewed_fields`** lists facts released from a held row before this fix, so the owner can review them. It's read-only and never lowers a tier.## [0.60.476.0] - 2026-10-27