Phase 4 & 5 validation report PDFs + consistent phase-series roadmaps - #137
Merged
Conversation
…indings renderer - check_provenance accepts a per-check 'source' override (top-level source is the default), so one synthesis report can pin numbers across several result JSONs. - _table_block forwards an optional highlight_color, so phase reports can use the series green instead of the default yellow. Backward compatible; existing prose and tests unchanged.
Complete the phase series with two prose-driven, provenance-checked reports: - Phase 4 (tool implementation code via iterative test-feedback repair): synthesizes the oracle-asymmetry campaign (12/20 = 0.60 deploy-reachable, corroborated 34/46 = 0.74; difficulty curve, leakage check, gate-gaming audit, evolver-vs-best-of-N). 27 numbers pinned across six committed source JSONs. - Phase 5 (continuous-improvement loop via a propose-only triage sentinel): leads with the shipped mechanism; the bug-supply and verification limits are carried honestly in a caveats section (cross-repo supply KILL pinned to source). Adds banked-artifact guards (check_provenance == [] for both new prose files) and a multi-source provenance unit test.
…ndex - Refresh the roadmap table in every phase report to one canonical, current view (each highlights its own phase). Correct the Phase 4 engine from the rejected 'Darwinian Evolver' (AGPL) to 'Iterative test-feedback repair' (MIT) and Phase 5 to 'Propose-only triage sentinel', in both the roadmap and the Approach engines table; re-render phases 1-3 from their run dirs (numbers unchanged). - Point the README phase table for Phases 4-5 at the new PDFs. - Add reports/README.md indexing the phase reports and findings deep dives. - Note the findings renderer in AGENTS.md.
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.
Completes the five-phase validation-report series. Phases 1-3 already shipped
phaseN_validation_report.pdf; Phases 4 and 5 were marked done in the README but linkedonly to markdown. This adds their PDFs in the same house style and makes the series
consistent.
What's here
reports/phase4_validation_report.pdf) — tool implementation code viaiterative test-feedback repair: the oracle-asymmetry campaign (12/20 = 0.60
deploy-reachable, corroborated 34/46 = 0.74; difficulty curve, leakage check,
gate-gaming audit, evolver-vs-best-of-N). 27 numbers pinned to six committed JSONs.
asymmetry_report.pdfstays as the deep dive.reports/phase5_validation_report.pdf) — the propose-only triage sentinel,framed capability-forward, with the bug-supply / verification limits carried honestly in
a caveats section (cross-repo supply KILL pinned to source).
generate_findings_report.pygains multi-source provenance (a check mayname its own
source) and ahighlight_colorpassthrough so the roadmap matches theseries green. Backward compatible.
own phase); corrected the Phase 4 engine from the rejected "Darwinian Evolver" (AGPL) to
"Iterative test-feedback repair" (MIT) and Phase 5 to "Propose-only triage sentinel" in
both the roadmap and the Approach engines table; re-rendered phases 1-3 from their run
dirs (measured numbers unchanged). README phase-table links updated; new
reports/README.mdindex; findings renderer noted in AGENTS.md.Tests
check_provenance == []for both new prose files, plus amulti-source provenance unit test.
pytest tests/ -q: 1665 passed, 1 skipped).