Repository navigation
review-kit 0.3.0: continue runs with previous, version every result - #17
Merged
Merged
Conversation
- previous: pass a run's continuation back to review its capped or blocked pairs, re-check its unverified findings and carry its confirmed ones in one run. Continuations chain and carry a digest; pairs: [] only re-verifies. Conflicting args stop the run. - Results carry version, and next when work is left; the first log line and the report header name the script version. A test keeps SCRIPT_VERSION in step with plugin.json. - The reporter gets compact JSON with nearby findings grouped as candidate duplicates, and combined coverage across runs. - Every run logs the report path and warns when it is the default at the repo root. - whenToUse shortened to 1,172 chars so the skill listing no longer cuts it; relaunch instructions moved into next. - README: Claude Code 2.1.305+, keep the host awake, stale-cache recovery, continuing a run. CI pins Claude Code 2.1.311. 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
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Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be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
previous: a run that leaves work returns acontinuation. Passing it back aspreviousreviews the capped or blocked pairs, re-checks unverified findings alongside them, carries confirmed findings, and reports combined coverage (e.g. "15 of 15 (12 in an earlier run)"). Continuations chain and carry a digest, so a mangled copy stops asbad-args.pairs: []only re-verifies.mode,since,changed,modules,lenses,verify,plan,carryandverifyOnlycannot be combined with it.versionon every result (failures too), in the first log line and in the report header. A test keepsSCRIPT_VERSIONequal toplugin.json.next: relaunch instructions now come with the result instead ofwhenToUse, which drops to 1,172 chars so the skill listing shows it whole (a test keeps it under 1,400).Test plan
node --test(182 pass)claude plugin validate --strictfor.,./review-kit,./token-audit,./token-meterclaude plugin test ./token-meter(40 pass)maxJobs: 2, thenresumeFromRunIdwithmaxJobs: 4; count cached agents and record in READMEwhenToUsein the listingscripts/seeded-defects-real-run.mjs)🤖 Generated with Claude Code