Repository navigation
Conversation
Anchors the ~2s/entry setup diagnosis and binding P1–P4 order so selection honesty, width-2, and native expansion cannot outrun amortized prep. Co-authored-by: Cursor <cursoragent@cursor.com>
Address review 47719: P1 gates width and broad native enrollment only, not five-minute step 3's bounded cohort. Fold P3 conjunction (P1 + index share / Rc→Arc) with the measured width-latch receipts, and bank the claim_batch retain-all lower bound from the #7597 50-entry harness. Co-authored-by: Cursor <cursoragent@cursor.com>
P1 must discriminate on the armed floor A/B receipt. Additive harness shares are a different denominator, unarmed ≠ Mode B, and #7597 forbids mechanism selection from those prices. Co-authored-by: Cursor <cursoragent@cursor.com>
Every discovery completion now prints selection_state, selected/total entry groups, ratio, and fallback_reason so whole-corpus affected runs are counted degradations, not quiet success. Co-authored-by: Cursor <cursoragent@cursor.com>
…x source Fixed 50-entry cohort run through the real floor path (run_discovery_corpus_with_options, Adaptive width) in two modes: A (production eviction on) and B (GUNBC_SCHEDULE_RETENTION_EVICT=0, retain-all pole). Entries 2..34 (Mode B OOM-killed at 34/50 after unbounded RSS growth, honestly reported and not backfilled) show statistically indistinguishable wall/resolve times between modes — disabling schedule-retention eviction does not collapse the ~1-2s per-entry tax. Verdict: REJECT the retention hypothesis; redirect to assembly/materialization reuse, per the parent's stated criterion. Also fixes the cargo fmt violation from the prior P1 instrumentation commit that was failing CI on this PR. Co-Authored-By: Claude Sonnet 5 <noreply@anthropic.com>
Publish selection_state, counts, ratio, and fallback_reason on the measurement tail, resolve receipt, floor-component JSON, and dedicated receipt file so carrier-only emission cannot satisfy the P2 bar. Co-authored-by: Cursor <cursoragent@cursor.com>
…out. Resolve floor-prep-tax-program.md conflicts with P1 BANKED (REJECT) status, P2 LANDED status, and e2e closeout routing to assembly/materialization. Co-authored-by: Cursor <cursoragent@cursor.com>
Contributor
6 tasks
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.
Auto-opened by session-dashboard for session
warm-lynx-428.Pushing to
session/warm-lynx-428advances this PR.Worker attestation
Before flipping this PR to ready for review, confirm each item:
npm test,cargo test) and the result.Closes #Ndirective.Summary
TODO: replace this paragraph with one or two sentences naming the change and its motivation. Reviewers read this first.
Test plan