Repository navigation
Conversation
States the multi-pass decision once, on measurements taken with search working and prompt v6, and records partitioning against the pass loop at equal session counts (behind on 7 of 8 pull requests at 2 and at 3 sessions). The superseded records carry a marker; code and workflow comments point at ADR-0024.
This branch was successfully deployed
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Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be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.
What this does
Adds ADR-0024, which supersedes ADR-0022, its addendum, and ADR-0023, so the multi-pass review decision is stated once.
SUPERSEDED by ADR-0024marker under the title. Nothing else in them changes.review_passes.py, the review workflow and the architecture diagram now cite ADR-0024.No behaviour changes.
The new measurement
8 public pull requests of 2,300-3,900 changed lines, 4 runs each, Opus 5.5 at high effort on prompt v6, blind judge reading source. Supported defect families per pull request:
Partitioning is behind on 7 of 8 pull requests at two and at three sessions (differences -1.1 and -0.6, both intervals include zero), and 15 to 30% cheaper. 351 of 353 partition findings were supported.
For review
The prose of
docs/adr/0024-a-review-is-up-to-four-passes-each-shown-what-came-before.md.🤖 Generated with Claude Code