Skip to content

Garry's Opinionated OpenClaw/Hermes Agent Brain

TypeScript

Stamp v0.60.698.0 and add Postgres arm weights for the GST-2515 and GST-2520 test files

garrytan9c1a46b3c349
diff --git a/.claude-plugin/plugin.json b/.claude-plugin/plugin.json
index 8c4b2da40e..a154f574e6 100644
--- a/.claude-plugin/plugin.json
+++ b/.claude-plugin/plugin.json
@@ -1,6 +1,6 @@
{
"name": "gbrain",
- "version": "0.60.697.0",
+ "version": "0.60.698.0",
"description": "Personal knowledge brain for your coding agent — hybrid search, synthesis, graph traversal, and durable cross-session memory over Postgres/PGLite with pgvector, plus a curated brain-first skill set.",
"author": {
"name": "Garry Tan",
diff --git a/.codex-plugin/plugin.json b/.codex-plugin/plugin.json
index 3dd5c5104c..bb868baa71 100644
--- a/.codex-plugin/plugin.json
+++ b/.codex-plugin/plugin.json
@@ -1,6 +1,6 @@
{
"name": "gbrain",
- "version": "0.60.697.0",
+ "version": "0.60.698.0",
"description": "Personal knowledge brain for your coding agent — hybrid search, synthesis, graph traversal, and durable cross-session memory over Postgres/PGLite with pgvector, plus a curated brain-first skill set.",
"author": {
"name": "Garry Tan",
diff --git a/BOOTSTRAP_FOR_AGENTS.md b/BOOTSTRAP_FOR_AGENTS.md
index 6982393d05..92c5261f53 100644
--- a/BOOTSTRAP_FOR_AGENTS.md
+++ b/BOOTSTRAP_FOR_AGENTS.md
@@ -1,4 +1,4 @@
-<!-- gbrain-runbook-stamp: 0.60.697.0 -->
+<!-- gbrain-runbook-stamp: 0.60.698.0 -->
<!-- This stamp must equal the VERSION file at every release; CI enforces it
(scripts/check-bootstrap-tag.sh). `gbrain bootstrap status` compares it to
the installed binary and warns on skew. -->
diff --git a/CHANGELOG.md b/CHANGELOG.md
index b559999a6c..c8044d45a9 100644
--- a/CHANGELOG.md
+++ b/CHANGELOG.md
@@ -10,7 +10,7 @@ credits are retained; no result has been reassigned to another provider. Origina
identifiers and attribution are available in the pre-removal Git revision
`6040075c6cb95be5881cc2e1b76ef7d71f4e5d29` (retained on 2026-09-23).
-## [Unreleased]
+## [0.60.698.0] - 2026-11-12
**A stalled background sync of a large source is retried with a budget that can finish it, and the retries no longer stop silently**
diff --git a/VERSION b/VERSION
index 4b12f1fec4..b55a94ab49 100644
--- a/VERSION
+++ b/VERSION
@@ -1 +1 @@
-0.60.697.0
+0.60.698.0
diff --git a/openclaw.plugin.json b/openclaw.plugin.json
index 1cf78c293f..e0c0ca74af 100644
--- a/openclaw.plugin.json
+++ b/openclaw.plugin.json
@@ -1,7 +1,7 @@
{
"id": "gbrain-context-engine",
"name": "gbrain",
- "version": "0.60.697.0",
+ "version": "0.60.698.0",
"description": "Personal knowledge brain with Postgres + pgvector hybrid search",
"family": "bundle-plugin",
"kind": "context-engine",
diff --git a/package.json b/package.json
index 93bacdc961..dc5a3498b7 100644
--- a/package.json
+++ b/package.json
@@ -255,7 +255,7 @@
"bun": ">=1.4.4"
},
"license": "MIT",
- "version": "0.60.697.0",
+ "version": "0.60.698.0",
"overrides": {
"@aws-sdk/util-format-url": "3.972.9",
"@ai-sdk/provider-utils": "4.0.56",
diff --git a/plugin-variants/gbrain-coding/.claude-plugin/plugin.json b/plugin-variants/gbrain-coding/.claude-plugin/plugin.json
index 24cbc85523..861d06804e 100644
--- a/plugin-variants/gbrain-coding/.claude-plugin/plugin.json
+++ b/plugin-variants/gbrain-coding/.claude-plugin/plugin.json
@@ -1,6 +1,6 @@
{
"name": "gbrain-coding",
- "version": "0.60.697.0",
+ "version": "0.60.698.0",
"description": "Brain-first coding agent working inside a repo: retrieval, routing, ingest discipline, correction hygiene. Default persona for the claude-code harness bridge; also published as the gbrain-coding marketplace variant. (persona variant of the gbrain plugin — 21 skills)",
"author": {
"name": "Garry Tan",
diff --git a/plugin-variants/gbrain-coding/.codex-plugin/plugin.json b/plugin-variants/gbrain-coding/.codex-plugin/plugin.json
index 86a8975b21..a3e1c98fca 100644
--- a/plugin-variants/gbrain-coding/.codex-plugin/plugin.json
+++ b/plugin-variants/gbrain-coding/.codex-plugin/plugin.json
@@ -1,6 +1,6 @@
{
"name": "gbrain-coding",
- "version": "0.60.697.0",
+ "version": "0.60.698.0",
"description": "Brain-first coding agent working inside a repo: retrieval, routing, ingest discipline, correction hygiene. Default persona for the claude-code harness bridge; also published as the gbrain-coding marketplace variant. (persona variant of the gbrain plugin — 21 skills)",
"author": {
"name": "Garry Tan",
diff --git a/plugin-variants/gbrain-coding/README.md b/plugin-variants/gbrain-coding/README.md
index 0888cc0cb7..d2624d05b6 100644
--- a/plugin-variants/gbrain-coding/README.md
+++ b/plugin-variants/gbrain-coding/README.md
@@ -1,4 +1,4 @@
-<!-- gbrain-plugin-tree-stamp: 0.60.697.0 -->
+<!-- gbrain-plugin-tree-stamp: 0.60.698.0 -->
# gbrain-coding (generated persona variant — do not hand-edit)
Brain-first coding agent working inside a repo: retrieval, routing, ingest discipline, correction hygiene. Default persona for the claude-code harness bridge; also published as the gbrain-coding marketplace variant.
diff --git a/plugin-variants/gbrain-daily/.claude-plugin/plugin.json b/plugin-variants/gbrain-daily/.claude-plugin/plugin.json
index 94e9b878ad..cf7a80de57 100644
--- a/plugin-variants/gbrain-daily/.claude-plugin/plugin.json
+++ b/plugin-variants/gbrain-daily/.claude-plugin/plugin.json
@@ -1,6 +1,6 @@
{
"name": "gbrain-daily",
- "version": "0.60.697.0",
+ "version": "0.60.698.0",
"description": "Personal knowledge-brain daily use: meetings, tasks, briefings, reading, research. Published as the gbrain-daily marketplace variant. (persona variant of the gbrain plugin — 20 skills)",
"author": {
"name": "Garry Tan",
diff --git a/plugin-variants/gbrain-daily/.codex-plugin/plugin.json b/plugin-variants/gbrain-daily/.codex-plugin/plugin.json
index f1489f9fa0..fda806963d 100644
--- a/plugin-variants/gbrain-daily/.codex-plugin/plugin.json
+++ b/plugin-variants/gbrain-daily/.codex-plugin/plugin.json
@@ -1,6 +1,6 @@
{
"name": "gbrain-daily",
- "version": "0.60.697.0",
+ "version": "0.60.698.0",
"description": "Personal knowledge-brain daily use: meetings, tasks, briefings, reading, research. Published as the gbrain-daily marketplace variant. (persona variant of the gbrain plugin — 20 skills)",
"author": {
"name": "Garry Tan",
diff --git a/plugin-variants/gbrain-daily/README.md b/plugin-variants/gbrain-daily/README.md
index 39e1961565..d781c843b0 100644
--- a/plugin-variants/gbrain-daily/README.md
+++ b/plugin-variants/gbrain-daily/README.md
@@ -1,4 +1,4 @@
-<!-- gbrain-plugin-tree-stamp: 0.60.697.0 -->
+<!-- gbrain-plugin-tree-stamp: 0.60.698.0 -->
# gbrain-daily (generated persona variant — do not hand-edit)
Personal knowledge-brain daily use: meetings, tasks, briefings, reading, research. Published as the gbrain-daily marketplace variant.
diff --git a/plugin/README.md b/plugin/README.md
index 499ea74996..1fd968075e 100644
--- a/plugin/README.md
+++ b/plugin/README.md
@@ -1,4 +1,4 @@
-<!-- gbrain-plugin-tree-stamp: 0.60.697.0 -->
+<!-- gbrain-plugin-tree-stamp: 0.60.698.0 -->
# gbrain plugin skill tree (generated — do not hand-edit)
This tree is the curated skill set for the gbrain Codex and Claude Code
diff --git a/scripts/postgres-arm-weights.json b/scripts/postgres-arm-weights.json
index 4decd48c5b..a6abbfd3f8 100644
--- a/scripts/postgres-arm-weights.json
+++ b/scripts/postgres-arm-weights.json
@@ -321,6 +321,7 @@
"test/migration-persistence-request-id-index.test.ts": 10916,
"test/migration-persistence-request-slug-index.test.ts": 13917,
"test/notifier-passed-freshness-marker-7941.test.ts": 31038,
+ "test/notifier-retry-unwatched-gst2515.test.ts": 7000,
"test/on-commit-embedding.serial.test.ts": 13575,
"test/oneshot-consumer-direct-route.test.ts": 6699,
"test/oneshot-housekeeping-cli.test.ts": 15812,
@@ -647,6 +648,7 @@
"test/sync-status-failure-detail-6513.test.ts": 7065,
"test/sync-status-truthful-7243.test.ts": 14660,
"test/sync-storage-invalid-8007.test.ts": 45745,
+ "test/sync-strict-skip-gst2520.test.ts": 19500,
"test/sync-strict-unmanaged-gst2492.test.ts": 16305,
"test/sync-tree-reuse-gbra133.test.ts": 16696,
"test/sync-unblock-repair-error-gbra130.test.ts": 9641,
diff --git a/src/core/behavior-change-notice.ts b/src/core/behavior-change-notice.ts
index 6b22fb379b..590ab065e6 100644
--- a/src/core/behavior-change-notice.ts
+++ b/src/core/behavior-change-notice.ts
@@ -1549,9 +1549,9 @@ export const BEHAVIOR_CHANGES: ReadonlyArray<{ since: string; text: ChangeText;
{ since: '0.60.686.0', id: 'readers_label_setter_gst2476', text: 'New `gbrain repair visibility (--slug <page>... | --slugs-file <file>) --readers <ids|none>` sets or removes a page\'s `readable_by` through the label-only door: the page keeps its trust tier and provenance and its file changes only on that line. Unknown or disabled members and unknown groups refuse `access_member_invalid`; a list without the brain owner refuses unless `--allow-without-owner`. Every `repair visibility` write to a managed page with no file outside a declared `storage.db_only` path is now held `canonical_file_missing` with nothing written, instead of failing `source_changed`.' },
{ since: '0.60.686.0', id: 'fence_fact_private_in_place_gst2487', text: 'A page write that only makes a `## Facts` row private (same claim, world to private) now makes the stored row private in place, keeping its id and trust tier, even when the row is more trusted than the writer. Before, such a row stayed live and world beside a contested private copy, so `gbrain repair visibility --class facts` could leave a fact remote callers still read while the page and its file read as fixed (`fact_left_readable`, `fact_left_world`). On a brain without managed persistence the repair now also makes the fenced row private. A write that makes a more trusted private row world is still held for the owner as before.' },
{ since: '0.60.688.0', id: 'sync_status_commits_waiting_gst2515', text: '`gbrain sync status` now reads `resumable` / `commits_waiting` when a managed source has nothing else pending (or only holds that need no person) but its checkout\'s HEAD is past the last synced commit. It used to read `synced` (`nothing_pending` or `held_by_design`) until some later sync ran. Its `next` is the plain sync (`gbrain sync` with `resume_argv`), which imports the waiting commits; status reads `synced` again once that sync lands. A script that treats `state: synced` as up to date now sees `resumable` while commits wait: run `next.argv`, or keep waiting for a resident to sync. A hold that needs a person (`held_needs_human`), a selection-scoped convergence, pending Git push holds and company-brain sources (which sync to their approved revision) keep their own reasons.' },
- { since: 'TBD', id: 'notifier_retry_unwatched_gst2515', text: 'The background syncer (the commit notifier in `gbrain serve`, `gbrain jobs work` or the autopilot daemon) now keeps retrying a source in `retry` / `notifier_stalled` on its backoff tier until a run completes, also with `sync.on_commit` off and across a restart or a change of the resident holding the notifier. Before, with `sync.on_commit` off, a source stopped being retried once its sync request was settled, while status still said it was retrying. A retry also no longer gets the 30-second first-run budget after a run that saved nothing: that budget now goes only to a source with no failed attempts on record, so a large source gets the full sync budget. `sync.on_commit` still defaults to off; this finishes runs a recorded request already asked for and does not start watching HEAD, and turning `sync.on_commit` off or excluding the source with `sync.on_commit.sources` still ends a retry that a HEAD move started.' },
- { since: 'TBD', id: 'strict_skip_mode_refusal_gst2520', text: '`gbrain sync --skip-failed`, and the automatic skip after repeated failed syncs, no longer move the checkpoint past files refused because of how the source or brain is set up (`strict_requires_managed`, a `canonical: git` source on a brain without managed persistence; `writer_coordinator_required`). The sync stays blocked and prints which setup to fix, and after the fix the next plain sync imports every page. A brain that was skipped past this way before upgrading needs one `gbrain sync --source <id> --full`. In a run with no such refusal, files that are themselves bad are skipped as before; a scheduled `gbrain sync --skip-failed && …` on a source in this state now stops at the blocked sync until its setup is fixed.' },
- { since: 'TBD', id: 'unmanaged_push_every_site_gst2456', text: 'Every push gbrain makes from a checkout that is not a managed worktree now scans the commits it publishes for secrets: `gbrain git commit-push` in commit mode and with `--pending` (it used to scan only with `--push-only --scan-secrets`), the post-commit hook `gbrain sources harden` installs and `scripts/brain-commit-push.sh`. A credential in an earlier local commit refuses `publish_secret_found` and nothing is pushed; the hook logs the refusal in brain-push.log. The helper no longer pushes on a machine without gbrain: it commits, logs the commit as local-only and exits 4. A scan that runs past its 300 s budget refuses `publish_scan_unreadable` naming the same command with `--scan-timeout <seconds>`, which `gbrain git commit-push`, `gbrain sources push` and `gbrain bootstrap repo` accept. No push follows tags (`push.followTags` is ignored). `gbrain sources push --json` prints a scan refusal (`blocked_history`, exit 5), and the access-policy commit reports one as `committed_not_pushed`.' },
+ { since: '0.60.698.0', id: 'notifier_retry_unwatched_gst2515', text: 'The background syncer (the commit notifier in `gbrain serve`, `gbrain jobs work` or the autopilot daemon) now keeps retrying a source in `retry` / `notifier_stalled` on its backoff tier until a run completes, also with `sync.on_commit` off and across a restart or a change of the resident holding the notifier. Before, with `sync.on_commit` off, a source stopped being retried once its sync request was settled, while status still said it was retrying. A retry also no longer gets the 30-second first-run budget after a run that saved nothing: that budget now goes only to a source with no failed attempts on record, so a large source gets the full sync budget. `sync.on_commit` still defaults to off; this finishes runs a recorded request already asked for and does not start watching HEAD, and turning `sync.on_commit` off or excluding the source with `sync.on_commit.sources` still ends a retry that a HEAD move started.' },
+ { since: '0.60.698.0', id: 'strict_skip_mode_refusal_gst2520', text: '`gbrain sync --skip-failed`, and the automatic skip after repeated failed syncs, no longer move the checkpoint past files refused because of how the source or brain is set up (`strict_requires_managed`, a `canonical: git` source on a brain without managed persistence; `writer_coordinator_required`). The sync stays blocked and prints which setup to fix, and after the fix the next plain sync imports every page. A brain that was skipped past this way before upgrading needs one `gbrain sync --source <id> --full`. In a run with no such refusal, files that are themselves bad are skipped as before; a scheduled `gbrain sync --skip-failed && …` on a source in this state now stops at the blocked sync until its setup is fixed.' },
+ { since: '0.60.698.0', id: 'unmanaged_push_every_site_gst2456', text: 'Every push gbrain makes from a checkout that is not a managed worktree now scans the commits it publishes for secrets: `gbrain git commit-push` in commit mode and with `--pending` (it used to scan only with `--push-only --scan-secrets`), the post-commit hook `gbrain sources harden` installs and `scripts/brain-commit-push.sh`. A credential in an earlier local commit refuses `publish_secret_found` and nothing is pushed; the hook logs the refusal in brain-push.log. The helper no longer pushes on a machine without gbrain: it commits, logs the commit as local-only and exits 4. A scan that runs past its 300 s budget refuses `publish_scan_unreadable` naming the same command with `--scan-timeout <seconds>`, which `gbrain git commit-push`, `gbrain sources push` and `gbrain bootstrap repo` accept. No push follows tags (`push.followTags` is ignored). `gbrain sources push --json` prints a scan refusal (`blocked_history`, exit 5), and the access-policy commit reports one as `committed_not_pushed`.' },
{ since: '0.60.691.0', id: 'commit_push_untrack_gst2519', text: 'New `gbrain git commit-push --untrack -m <msg> -- <paths>` stops tracking files and keeps them on disk: it removes every file the last commit holds under each named path from Git and the index and commits only those deletions, under the commit lock (`--dry-run --untrack` previews and refuses the same way). In an ordinary checkout the commit is pushed. In a managed canonical worktree it only takes files under a `storage.db_only` directory that `gbrain.yml` at HEAD declares and `.gitignore` ignores, refusing anything else `untrack_not_db_only` with the index as it was, and never pushes; the pages stay live. A plain commit-push of a removal staged by hand with `git rm --cached` now names the `--untrack` command instead of blaming a nested repository. Members under `GBRAIN_ENFORCED_MEMBER` are refused `--untrack`.' },
{ since: '0.60.691.0', id: 'db_only_labels_kept_gst2519', text: 'A page under a declared `storage.db_only` directory whose file Git still tracks keeps its privacy labels. On a git-canonical (strict) source the next sync no longer imports the unchanged committed file over a database change such as a `gbrain repair visibility` label repair, which used to undo the repair and widen the page again. On every managed source a commit to such a file no longer widens the page: a `visibility`, `access`, `confidentiality` or `readable_by` label the page has and the file does not state is kept, and a label the file states applies, a wider one included. Files outside `storage.db_only` import as they stand.' },
{ since: '0.60.691.0', id: 'git_ignored_label_repair_gst2519', text: 'On a git-canonical (strict) source, `gbrain repair visibility` on a page stamped `git_ignored` (saved under a path the checkout\'s Git ignores, so it has no file) now writes the label in the database, where the page lives, instead of holding it `canonical_file_missing`. A page whose ignore rule is gone is still held.' },
diff --git a/templates/bootstrap/template-repo/README.md b/templates/bootstrap/template-repo/README.md
index 6ea344ee67..a7a16fca75 100644
--- a/templates/bootstrap/template-repo/README.md
+++ b/templates/bootstrap/template-repo/README.md
@@ -1,6 +1,6 @@
# gbrain agent workspace — template
-<!-- gbrain-template-stamp: 0.60.697.0 -->
+<!-- gbrain-template-stamp: 0.60.698.0 -->
This repository is the **"Use this template"** distribution artifact for a
[gbrain](https://github.com/garrytan/gbrain) personal-agent workspace — the same