Find the work behind the code.

GH issue #181: https://github.com/unbraind/pm-gantt-chart/issues/181 · author @github-actions[bot] · state reason completed · created 2026-10-25T09:40:42Z · updated 2026-10-25T23:08:47Z

GH issue #182: https://github.com/unbraind/pm-gantt-chart/issues/182 · author @unbraind · state reason completed · created 2026-10-25T21:54:33Z · updated 2026-10-26T08:22:19Z

GH issue #187: https://github.com/unbraind/pm-gantt-chart/issues/187 · author @github-actions[bot] · state reason completed · created 2026-10-26T10:48:59Z · updated 2026-10-26T14:42:46Z

GH issue #193: https://github.com/unbraind/pm-gantt-chart/issues/193 · author @github-actions[bot] · state reason completed · created 2026-10-27T10:25:28Z · updated 2026-11-02T09:26:38Z

GH issue #211: https://github.com/unbraind/pm-gantt-chart/issues/211 · author @github-actions[bot] · created 2026-11-03T16:26:22Z · updated 2026-11-05T12:05:20Z

GH issue #214: https://github.com/unbraind/pm-gantt-chart/issues/214 · author @unbraind · state reason completed · created 2026-11-05T15:16:09Z · updated 2026-11-07T02:37:33Z

GH issue #159: https://github.com/unbraind/pm-slack/issues/159 · author @unbraind · state reason completed · created 2026-10-25T21:55:40Z · updated 2026-10-26T08:22:56Z

Investigate release CI checkpoint promotion lock conflict from GitHub issue #152; characterize loaded contention, preserve declared bounds and checkpoint roots, and prove any correction by behavioral revert.

Investigate repeatable local acceptance latency without changing authored limits or coverage scope. In the PR 145 review run, test/process-cli.test.ts reproduced spawnSync ETIMEDOUT alone under its existing command limit

The full release gate observed an existing packed LM CLI timing failure: measured CLI 22138 ms against its unchanged 20000 ms bound. The same unmodified test passed alone at 8955 ms. Investigate repeatable runtime and IO

Follow up pm-qz1k by reconciling the saved typed-binding review against main. PR #191 already delivered all saved source and test hunks. Add independent end-to-end regression coverage and isolated behavioral revert proof

Implement pm-ops#249: durable candidate intent, hourly reconciliation and daily catch-up, protected auto-merge, owned supersession and the local caller pilot.

Strict pm health passes but reports an existing stale in-progress item and legacy provenance role values. Review owner activity and provenance normalization separately from PR #62 lease recovery; preserve audit history a

Same-inode rename plus edit and unambiguous unchanged-content moves retain FileIds today. An edited move across filesystems can lose both inode and content evidence, so guessing can conflate independent files. Add an exp

Replace generated standalone CLI direct codec dependency with public SDK codecs; preserve certified read-only receipts and document the supported host contract.

Current adoption scope: the November 6 watch covers 21 failing sync runs and 15 failing closeout runs across 22 public callers. Start with the complete per-caller matrix and required actions in docs/scheduled-workflow-wa

The November 6 fleet watch still samples old workflow releases with silent exact-version verification failures and hidden closeout stderr. Existing normalization, pagination and no-op fixes require caller pin adoption. S

GH issue #280: https://github.com/unbraind/pm-web/issues/280 · author @unbraind · state reason not_planned · created 2026-10-29T19:07:13Z · updated 2026-11-06T19:47:10Z

Issue 197 remains open after main 300e115 and PR 223. The finalizer appends an independent import event from the committing observer; ownership-guarded SDK compensation refuses it after an ordinary failure. Native reprod

Public caller replays show pm-changelog and pm-ops gated atomic dry runs returning a plan while the identical local apply refuses configured create or completeness rules. The importer checks completeness, identity and pr