Repository navigation
feat(scorecard): session log, memory handles, settings and the card renderer - #27
Merged
Merged
Conversation
A push publishes every commit it carries, and a blob stays fetchable by SHA after a later commit deletes it, so the pre-push gate now scans each new commit, not only the tip. The tip gets the whole-tree scan; earlier commits are scanned on the files they touched. Extensionless dotfiles such as .gitignore are text too. The self-test plants both cases. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
One append-only log per host session (scorecard/<session_id>.jsonl) that every hook appends to and the scorecard folds at render time, plus short hashed handles ([mxxxx]) so injected memory lines can be credited by id. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
sessionScore (off|compact|full) controls the per-turn memory scorecard; outcomeReview (off|nudge|stop) controls how hard Claude is asked to credit memory by id. Codex defaults to off and nudge. Documents both, the scorecard's definitions and the review's cost in the user guides and option tables. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
A pure fold over the session log gives per-lesson shown/used/unknown and the worked/failed/waiting verdict (explicit verdict, correction, last failed non-read-only tool call, next prompt), and a renderer prints the exact card. Golden fixtures pin the summaries and the text. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
4 tasks done
4 tasks done
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
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.
Summary
First of five PRs that add the session memory scorecard. Nothing calls this code yet, so behaviour does not change.
<dataDir>/scorecard/<session_id>.jsonl, that the hooks will write to.[m7k2q]for memory entries, derived from the entry id, so the model can name an entry cheaply.sessionScore(off|compact|full) andoutcomeReview(off|nudge|stop).What to review
That the counts always add up: shown = used + not used + unknown, and used = worked + failed + waiting.
Merge order
Merge after #26. This PR targets
pre-maindirectly, so until #26 merges its diff also shows the earlier PRs in the series.Test plan
integrations/claude-code:npm run lint,npm test,npm run test:dist,node scripts/verify-manifests.mjsintegrations/codex:npm test,npm run test:distnode .github/scripts/leakcheck.mjs --strict, and the pre-push scan over every commit in this PR🤖 Generated with Claude Code