docs(napkin-math): refresh methodology + plan status for 2026-05-22 ship-set#754
Merged
Conversation
…hip-set Updates two docs to reflect the post-#753 state of the napkin-math pipeline. methology.md: describe the current pipeline behaviour — two-batch compress with paraphrase-tolerant quote match and cross-bucket promoter; extract's source-arithmetic preservation, threshold-pairing, and dropped_signals field; 19-check validator (added aggregate_not_bounded, requirement_has_margin, dropped_signals_schema); bounds' asymmetric source label on commitment defaults, calculation-output strip, reserved correlations block, reserved lognormal/pert disciplines with loud NotImplementedError; advisory audit_source_preservation.py step. 20260520_plan.md → 20260522_plan.md: bump status date; mark PR #750 merged; add PR #751/#752/#753 entries (proposal 141 implementation); update Phase status table (added 4.5 audit row, reclassified Phase 8 as partially done, Phase 10 marked done for current ship-set); add v58 14-plan empirical snapshot (1 viable / 5 fragile / 8 doom); reorder Next likely move now that proposal 141 has shipped — Phase 5 citation verifier promoted to #1, Phase 8 samplers added as #2 with v58 cases that bite now, Phase 9 composite-band cap as #3. Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
Three review fixes: 1. plan: update the stale 'No formal source-preservation audit implementation' bullet — Fork B shipped in PR #751/#752/#753; Fork A, orchestrator-side prior-baseline injection, and strict-mode are the actual still-pending follow-ups. 2. plan: bump the document title from 2026-05-20 to 2026-05-22; add an italicised note that the doc was originally drafted 2026-05-20 and renamed/refreshed for the post-#753 ship-set. 3. methology: stop overclaiming what the assessment Basis column exposes. summarize_assessment.py maps source:'data' → 'report_derived' and source:'assumption' → 'model_assumption', and that is what the column shows; the finer 'plan-internal gap forecast vs bare commitment' distinction lives in the rationale string, not the column. Co-Authored-By: Claude Opus 4.7 (1M context) <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
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
methology.md: rewritten to describe the current pipeline behaviour — two-batch compress with paraphrase-tolerant quote match and cross-bucket promoter; extract's source-arithmetic preservation, threshold-pairing, anddropped_signalsfield; 19-check validator (addedaggregate_not_bounded,requirement_has_margin,dropped_signals_schema); bounds' asymmetric source label on commitment defaults, calculation-output strip, reservedcorrelationsblock, reservedlognormal/pertdisciplines with loudNotImplementedError; advisoryaudit_source_preservation.pystep.20260520_plan.md→20260522_plan.md: bumped status date; marked PR napkin-math(compress): cross-bucket promoter for gate-shaped items misfiled under risks #750 merged; added PR napkin-math: advisory source-preservation audit (Fork B, proposal 141 PR 1) #751/napkin-math: dropped_signals schema + validator + audit consumption (proposal 141 PR 2) #752/napkin-math: prior-signal ledger orchestration for extract skill (proposal 141 PR 3) #753 entries (proposal 141 implementation); updated Phase status table (added 4.5 audit row, reclassified Phase 8 as partially done, Phase 10 marked done for current ship-set); added a v58 14-plan empirical snapshot (1 viable / 5 fragile / 8 doom); reordered "Next likely move" now that proposal 141 has shipped — Phase 5 citation verifier promoted to fix: Add UTF-8 encoding for markdown file operations #1, Phase 8 samplers added as Ath dev #2 with v58 cases that bite now, Phase 9 composite-band cap as Add support for HuggingFace GGUF models in Ollama #3.Docs-only PR. No code, no prompts, no schema changes.
Test plan
methology.mdon GitHub — section headings, sampling-discipline bullet list, and theaudit_source_preservation.pyparagraph all read cleanly.20260522_plan.mdon GitHub — confirm the Phase status table formats correctly with the new 4.5 row, and the v58 verdict table renders.grep -rn "20260520_plan" .returns no stale references (the rename was hunted for callers).🤖 Generated with Claude Code