chore(agent-memory): upgrade 4.42.1 → 4.42.2 — the step-6 subject read starts with the window's commits - #108
Merged
Merged
Conversation
…d starts with the window's commits Mode B upgrade (PATCH, guidance only). Reconcile re-copied DECAY.md and REVIEW.md (stock 4.42.1 before); the protocol is unchanged, so there is no semantic step. Stamped 4.42.2. Ran in a throwaway worktree of origin/main. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
1 participant
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
What
agent-memory Mode B upgrade 4.42.1 → 4.42.2 (PATCH, guidance only). Reconcile re-copied
DECAY.mdandREVIEW.md— each was stock 4.42.1, so staleness only. The protocol, hooks, the CI floor, the schema and the built-ins are unchanged (no semantic step). Stamped 4.42.2; both re-copied files byte-identical to their MANIFEST sources;memory-lint: 0 errors, 1 warning(s) — one completed thread pastarchive_window, for the next review's step-5 sweep (Python/Node identical);memory-lint --stagedon this change → ok. Ran in a throwaway git worktree oforigin/main; the checkout was not touched.What changes for this repo's reviews.
REVIEW.mdstep 6's declaration-gap read now looks in two places, commits first:git log --since=<oldest window log's date> -- <paths the fact names>(leaving out hub paths that most commits touch — the measured cut keeps a path touched by at most 5% of all commits), then the session logs as before. Every hit is held to the decision test — did the session rely on what the fact states? — and a fact recording an event (a release shipped, work completed) is not kept alive by later work on the same code.Why
v4.42.1's read searched only the window's session logs. Measured on the tool repo's own history (RFC-0006), that search found neither of the two sessions behind its one wrongful fact archival — sessions describe their work in prose, not by the paths a fact names — while the window's commits touching those paths found both, with no flag on any correct archival once hub paths are left out. The decision-test sentence stops the check from keeping release records alive whenever the code they describe evolves. Upstream release: https://github.com/Accenture/mercury-go/releases/tag/v4.42.2
Co-Authored-By: Claude Opus 5.5 noreply@anthropic.com
🤖 Generated with Claude Code