[PR #2701] docs(skills): add source attribution and evidence handoff guidance #5324

New issue

Open

opened 2026-10-31 01:05:44 +00:00 by bsakaguchi · 0 comments

bsakaguchi commented

2026-10-31 01:05:44 +00:00

Owner

📋 Pull Request Information

Original PR: https://github.com/morluto/rea/pull/2701
Author: @rohitg00
Created: 10/13/2026
Status: 🔄 Open

Base: main ← Head: fix/composite-cleanup-diagnostics

📝 Commits (10+)

2b0b65d fix(binary): retain all composite cleanup failures
611394e chore: integrate upstream CI fixes
7be3492 chore: integrate upstream CI test correction
443b404 test(apple): bound cold archive fixture compilation
79ba2a8 chore: merge upstream CI and provenance test fixes
e439438 test(release): scope preflight schema to selected jobs
97ec5a5 test(ci): include cold-cache Apple fixture timeout correction
c417687 test(ci): merge shared release and Apple fixture fixes
b62abb9 chore: merge upstream CI fixes into cleanup branch
3bed7cd Merge remote-tracking branch 'origin/main' into fix/composite-cleanup-diagnostics

📊 Changes

6 files changed (+253 additions, -11 deletions)

View changed files

📝 .agents/skills/reverse-engineer-anything/SKILL.md (+7 -4)
📝 .agents/skills/reverse-engineer-anything/references/evidence-workflows.md (+95 -6)
➕ .agents/skills/reverse-engineer-anything/references/web-feature-attribution.md (+147 -0)
📝 src/application/SetupSkill.ts (+1 -0)
📝 src/generatedPackageMetadata.ts (+1 -1)
📝 tests/boundary/filesystem/setupSkill.test.ts (+2 -0)

📄 Description

Summary

Add an installed skill workflow for attributing captured website code and handing off an investigation without losing Evidence provenance or unresolved questions. Agents can select retained script bytes, interpret source-map coordinates and module candidates, and resume from a portable Evidence bundle while preserving coverage limits.

Change and scope

Add the authored web-feature-attribution.md reference for script export, source-map lookup and module attribution. Keep source identity, execution coverage and feature causality distinct, including the scenario response-body limitation.
Extend Evidence guidance with discriminating probes, residual-unknown revision handling, export/import handoffs and historical-observation limits. Raw error observations without recorded Evidence IDs require separate preservation.
Register the reference in the managed skill inventory, update its two installation expectations, and advance skill metadata to version 43.

This is the remaining skills contribution requested in the maintainer's scope review. Already-delivered cleanup and retention changes are removed from the diff against main. Other runtime additions, including Evidence-summary pagination, are excluded from this narrowed PR; the guidance uses the current unpaged summary contract.

Contract and boundary impact

The authored .agents skill sources own the guidance. Managed installation includes the new reference; generated skills/ output is not committed. CLI/MCP schemas, analysis implementations, provider behavior and session semantics match main. No dependency, migration or new execution authority is introduced.

Documentation or usability only; no target interaction.

Evidence and validation

Reviewed all six changed files against main bec88b9d and traced the guidance to current source contracts and documentation. Applied the code-simplifier review to the small inventory/metadata changes.
oxfmt --check on all six changed files: passed.
git diff --check origin/main: passed.
Source audit: five fenced JSON examples parse; 27 linked repository source paths exist; skill version and managed reference entries agree. Package metadata regenerated with the existing generator.
Updated installation expectations were inspected, not executed. Tests, builds, packaged installation and provider runs remain skipped at the maintainer's request. Earlier test and CI receipts do not validate this revision. Live workflow behavior remains unverified.

Author readiness

Human owner: @rohitg00

I reviewed the final diff and relevant existing code, understand the change and its edge cases, and take responsibility for the complete PR, including AI-assisted work.
I checked the evidence above, confirmed it supports the current change, and recorded meaningful validation gaps.
I declare this PR ready for initial maintainer review.

_{🔄 This issue represents a GitHub Pull Request. It cannot be merged through Gitea due to API limitations.}

## 📋 Pull Request Information **Original PR:** https://github.com/morluto/rea/pull/2701 **Author:** [@rohitg00](https://github.com/rohitg00) **Created:** 10/13/2026 **Status:** 🔄 Open **Base:** `main` ← **Head:** `fix/composite-cleanup-diagnostics` --- ### 📝 Commits (10+) - [`2b0b65d`](https://github.com/morluto/rea/commit/2b0b65dc583e905c322df050cbf2987588fff102) fix(binary): retain all composite cleanup failures - [`611394e`](https://github.com/morluto/rea/commit/611394e47f1df0a91f3fe64cc5c260b901cb29a4) chore: integrate upstream CI fixes - [`7be3492`](https://github.com/morluto/rea/commit/7be34929d9f18d0fe235c5245ef67fa539ac157e) chore: integrate upstream CI test correction - [`443b404`](https://github.com/morluto/rea/commit/443b404caed1391f5348e4b5a21cc586b49bbb8c) test(apple): bound cold archive fixture compilation - [`79ba2a8`](https://github.com/morluto/rea/commit/79ba2a89d7d0a414e4f41b6bbd721e4d2a1cdec5) chore: merge upstream CI and provenance test fixes - [`e439438`](https://github.com/morluto/rea/commit/e4394384a97c5d7cd0d560b546786965a9c283b5) test(release): scope preflight schema to selected jobs - [`97ec5a5`](https://github.com/morluto/rea/commit/97ec5a57f191018a9e581cea1fe747da6ba39d86) test(ci): include cold-cache Apple fixture timeout correction - [`c417687`](https://github.com/morluto/rea/commit/c417687c0d6ce8f6e9d9a002c12dd7195c705de7) test(ci): merge shared release and Apple fixture fixes - [`b62abb9`](https://github.com/morluto/rea/commit/b62abb968073fc037847141d2f726b0da54dcfae) chore: merge upstream CI fixes into cleanup branch - [`3bed7cd`](https://github.com/morluto/rea/commit/3bed7cd0ee198b344969efbe3fe07f0ee3b90a28) Merge remote-tracking branch 'origin/main' into fix/composite-cleanup-diagnostics ### 📊 Changes **6 files changed** (+253 additions, -11 deletions) <details> <summary>View changed files</summary> 📝 `.agents/skills/reverse-engineer-anything/SKILL.md` (+7 -4) 📝 `.agents/skills/reverse-engineer-anything/references/evidence-workflows.md` (+95 -6) ➕ `.agents/skills/reverse-engineer-anything/references/web-feature-attribution.md` (+147 -0) 📝 `src/application/SetupSkill.ts` (+1 -0) 📝 `src/generatedPackageMetadata.ts` (+1 -1) 📝 `tests/boundary/filesystem/setupSkill.test.ts` (+2 -0) </details> ### 📄 Description ## Summary Add an installed skill workflow for attributing captured website code and handing off an investigation without losing Evidence provenance or unresolved questions. Agents can select retained script bytes, interpret source-map coordinates and module candidates, and resume from a portable Evidence bundle while preserving coverage limits. ## Change and scope - Add the authored `web-feature-attribution.md` reference for script export, source-map lookup and module attribution. Keep source identity, execution coverage and feature causality distinct, including the scenario response-body limitation. - Extend Evidence guidance with discriminating probes, residual-unknown revision handling, export/import handoffs and historical-observation limits. Raw error observations without recorded Evidence IDs require separate preservation. - Register the reference in the managed skill inventory, update its two installation expectations, and advance skill metadata to version 43. This is the remaining skills contribution requested in the [maintainer's scope review](https://github.com/morluto/rea/pull/2701#issuecomment-6186265679). Already-delivered cleanup and retention changes are removed from the diff against main. Other runtime additions, including Evidence-summary pagination, are excluded from this narrowed PR; the guidance uses the current unpaged summary contract. ## Contract and boundary impact The authored `.agents` skill sources own the guidance. Managed installation includes the new reference; generated `skills/` output is not committed. CLI/MCP schemas, analysis implementations, provider behavior and session semantics match main. No dependency, migration or new execution authority is introduced. - [x] Documentation or usability only; no target interaction. ## Evidence and validation - Reviewed all six changed files against main `bec88b9d` and traced the guidance to current source contracts and documentation. Applied the code-simplifier review to the small inventory/metadata changes. - `oxfmt --check` on all six changed files: passed. - `git diff --check origin/main`: passed. - Source audit: five fenced JSON examples parse; 27 linked repository source paths exist; skill version and managed reference entries agree. Package metadata regenerated with the existing generator. - Updated installation expectations were inspected, not executed. Tests, builds, packaged installation and provider runs remain skipped at the maintainer's request. Earlier test and CI receipts do not validate this revision. Live workflow behavior remains unverified. ## Author readiness Human owner: @rohitg00 - [x] I reviewed the final diff and relevant existing code, understand the change and its edge cases, and take responsibility for the complete PR, including AI-assisted work. - [x] I checked the evidence above, confirmed it supports the current change, and recorded meaningful validation gaps. - [x] I declare this PR ready for initial maintainer review. --- <sub>🔄 This issue represents a GitHub Pull Request. It cannot be merged through Gitea due to API limitations.</sub>

bsakaguchi added the

pull-request

label

2026-10-31 01:05:44 +00:00

bsakaguchi referenced this issue

2026-11-01 21:43:47 +00:00

[PR #5324] fix(javascript-recovery): run Wakaru on macOS arm64 with shell resource limits #6880