Qualify branch-aware Go retrieval and source views - #25
Merged
Merged
Conversation
ashwingopalsamy
enabled auto-merge (squash)
September 27, 2026 07:56
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Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be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.
Summary
agentic-go source-viewto materialize a selected branch at an exact commit in a visible detached worktree. It defaults to localmain(then configuredorigin/HEAD), supports explicit refs, and only carries a dirty overlay when the source checkout is at the exact selected commit.Retrieval evidence and decision
The corrected medium screen reached Recall@10 0.35 and Precision@10 0.20, with no document retrieval. The fresh four-question text-candidate ablation recovered all 15 reviewed gold spans in complete candidate pools, but the mixed top-10 ranking reached macro Recall 0.1625, Precision 0.075, and MRR 0.28125. Small-repository relevance was unusable; Kubernetes source coverage was partial and its one-run warm p95 was 10.52 seconds. Process RSS was not measured. The fixed
rgscorer and optional gopls arm are not a competent, complete native-agent baseline.These results do not establish useful agent outcomes, retrieval superiority, arbitrary-scale support, or a persistence bottleneck. The text ablation is evaluation-only. This PR adds no persistent index and no MCP branch selector. The source-view command is a branch-scoped setup preview. A future matched engineering screen remains gated on useful retrieval and a competent native baseline. The six-cell selector reliability screen is separate: safety passed, efficiency promotion failed; the stale 18-cell rerun was not run.
New model-based council/review work used GPT-6 Luna Max only. Historical GPT-6 Sol findings remain separately labeled.
Verification
go test ./...go test -race ./...go vet ./...go build ./...git diff --check 7b5111c6365a2a12806a561d86745d5bc50d7c9e..HEADAll listed checks passed locally. No release or tag was created.