feat(me): migrate autoresearch into me plugin with /goal integration - #660
Merged
Conversation
added 8 commits
May 26, 2026 08:57
This reverts commit a5a1854.
Autonomous experiment loop migrated from plugins/autoresearch/ to me plugin. Termination owned by Claude Code /goal command instead of self-managed 'NEVER STOP' policy. JSONL state protocol, dashboard, worklog, ideas backlog preserved.
Remove standalone plugins/autoresearch/ and its marketplace entries. The skill now lives at plugins/me/skills/autoresearch/SKILL.md (added in the previous commit) with /goal-driven termination instead of the former 'NEVER STOP' self-managed policy. Also update CI workflow, codex sync check script, and BATS expectations that hardcoded references to plugins/autoresearch. Note: marketplace entry removal and plugin directory deletion are combined into one commit because the BATS test suite verifies bidirectional parity between marketplace.json and plugins/ directory -- they must change atomically.
- Remove autoresearch@bstack from enabledPlugins in setup settings (standalone plugin no longer exists) - Drop the autoresearch directory tree entry and dedicated plugin section from README - Add autoresearch row to the me plugin skill table since the skill now lives there
main's revert (c065e08) re-introduced plugins/autoresearch/ and related references during the merge. This commit re-removes them to preserve the migration to plugins/me/skills/autoresearch/.
📝 WalkthroughWalkthroughThis PR completes the migration of the autoresearch autonomous experiment loop from a standalone plugin to an integrated skill within the me plugin. It replaces the prior "never stop" infinite-loop termination model with goal-driven per-turn evaluation, removes legacy plugin artifacts and marketplace entries, and updates all validation tests and documentation accordingly. ChangesAutoresearch Plugin to Skill Integration
Estimated Code Review Effort🎯 3 (Moderate) | ⏱️ ~20 minutes Possibly Related PRs
Poem
🚥 Pre-merge checks | ✅ 4 | ❌ 1❌ Failed checks (1 warning)
✅ Passed checks (4 passed)
✏️ Tip: You can configure your own custom pre-merge checks in the settings. ✨ Finishing Touches📝 Generate docstrings
🧪 Generate unit tests (beta)
Thanks for using CodeRabbit! It's free for OSS, and your support helps us grow. If you like it, consider giving us a shout-out. Comment |
![coderabbitai[bot]](https://avatars.githubusercontent.com/in/347564?s=60&v=4){kind=small}
There was a problem hiding this comment.
Actionable comments posted: 2
Caution
Some comments are outside the diff and can’t be posted inline due to platform limitations.
⚠️ Outside diff range comments (1)
README.md (1)
49-49:⚠️ Potential issue | 🟡 Minor | ⚡ Quick winFix README skill count for the
meplugin.
- README.md line 49 says “Personal skills (16 total)”, but the
meskills table lists only 13 skills (includingautoresearch).- The
plugins/me/skills/directory contains many more skill folders (22 excludingsetup), so the “16 total” figure doesn’t match the table or the actual skill set.- Update the count and/or the table contents so all three are consistent.
🤖 Prompt for AI Agents
Verify each finding against current code. Fix only still-valid issues, skip the rest with a brief reason, keep changes minimal, and validate. In `@README.md` at line 49, Update the README so the "Personal skills (16 total)" entry for the me plugin matches the actual skills: either (A) update the count to the real number of skill folders found under plugins/me/skills/ (22 excluding setup) and ensure the skills table lists all 22 names (including autoresearch), or (B) prune the table to exactly match 16 skills if you intend to advertise 16 — make the README table, the numeric total, and the actual plugins/me/skills/ directory contents consistent; modify the skills table in README.md and the total label so all three sources (count, table, directory) agree.
🧹 Nitpick comments (4)
docs/superpowers/specs/2026-05-26-autoresearch-me-integration-design.md (3)
144-158: 💤 Low valueSpecify language for fenced code block.
The usage example should specify a language identifier (use
textor leave as-is if demonstrating console interaction).♻️ Proposed fix
### 동작 예시 -``` +```text User: /me:autoresearch "optimize parser perf, R² > 0.85" 또는 User: "parser 성능을 R² 0.85까지 끌어올리는 실험 루프 돌려줘"🤖 Prompt for AI Agents
Verify each finding against current code. Fix only still-valid issues, skip the rest with a brief reason, keep changes minimal, and validate. In `@docs/superpowers/specs/2026-05-26-autoresearch-me-integration-design.md` around lines 144 - 158, The fenced example block lacks a language identifier; update the code fence that opens before "User: /me:autoresearch ..." to use a language tag (e.g., ```text) so the snippet is rendered correctly; locate the fenced block containing the "User: /me:autoresearch \"optimize parser perf, R² > 0.85\"" example and change the opening triple backticks to "```text" (and keep the closing triple backticks) to follow the proposed fix.
35-40: 💤 Low valueSpecify language for fenced code block.
The directory structure example should specify a language identifier (or use
textif no syntax highlighting is needed).♻️ Proposed fix
### 결과 디렉토리 구조 -``` +```text plugins/me/skills/autoresearch/ └── SKILL.md # 단일 파일 plugins/autoresearch/ # 삭제</details> <details> <summary>🤖 Prompt for AI Agents</summary>Verify each finding against current code. Fix only still-valid issues, skip the
rest with a brief reason, keep changes minimal, and validate.In
@docs/superpowers/specs/2026-05-26-autoresearch-me-integration-design.md
around lines 35 - 40, The fenced directory-structure code block in the docs
containing "plugins/me/skills/autoresearch/" lacks a language identifier; update
the opening fence to include a language (e.g., changetotext) so the
block is rendered without erroneous syntax highlighting—edit the block that
contains the lines "plugins/me/skills/autoresearch/", "└── SKILL.md", and
"plugins/autoresearch/ # 삭제" to use ```text as the fence.</details> --- `62-111`: _💤 Low value_ **Specify language for fenced code block.** The structure outline should specify a language identifier (use `text` or `markdown`). <details> <summary>♻️ Proposed fix</summary> ```diff ### SKILL.md 본문 구조 -``` +```text # autoresearch: Autonomous experiment loop ## When to use ``` </details> <details> <summary>🤖 Prompt for AI Agents</summary>Verify each finding against current code. Fix only still-valid issues, skip the
rest with a brief reason, keep changes minimal, and validate.In
@docs/superpowers/specs/2026-05-26-autoresearch-me-integration-design.md
around lines 62 - 111, The top-level fenced code block that contains the
autoresearch spec currently lacks a language identifier; update the opening
fence (thethat precedes the line "# autoresearch: Autonomous experiment loop") to include a language tag such as text or markdown (e.g., changetoplaintext styling for the whole block.docs/superpowers/plans/2026-05-26-autoresearch-me-integration.md (1)
477-489: 💤 Low valueAdd blank lines around fenced code block.
The expected output example needs blank lines before and after the code fence, and should specify a language identifier for the fence.
♻️ Proposed fix
Expected: +
name: autoresearch
description: Use when asked to "run autoresearch", ...
allowed-tools:
- Read
- Write
- Edit
- Bash
- Glob
- Grep
+🤖 Prompt for AI Agents
Verify each finding against current code. Fix only still-valid issues, skip the rest with a brief reason, keep changes minimal, and validate. In `@docs/superpowers/plans/2026-05-26-autoresearch-me-integration.md` around lines 477 - 489, Add blank lines before and after the fenced code block and include a language identifier on the opening fence; specifically, ensure there is an empty line above the triple-backtick fence that contains the YAML starting with "name: autoresearch" and an empty line after the closing fence, and change the opening fence to something like "```yaml" so the block is properly recognized as YAML.
🤖 Prompt for all review comments with AI agents
Verify each finding against current code. Fix only still-valid issues, skip the
rest with a brief reason, keep changes minimal, and validate.
Inline comments:
In `@docs/superpowers/plans/2026-05-26-autoresearch-me-integration.md`:
- Line 516: The checklist entry "[ ] SKILL.md가 200줄 이하" conflicts with the
actual SKILL.md (239 lines); either update the criterion text to match the real
file (e.g., change the checkbox label to "SKILL.md가 240줄 이하" or "≤240 lines") or
reduce the SKILL.md content to <=200 lines; locate the checklist line containing
"[ ] SKILL.md가 200줄 이하" and either edit that label to the new limit or refactor
the file named SKILL.md to remove/simplify sections until its line count is
<=200, ensuring the checkbox and the actual file length are consistent.
In `@docs/superpowers/specs/2026-05-26-autoresearch-me-integration-design.md`:
- Line 178: The success criterion in
2026-05-26-autoresearch-me-integration-design.md that states "SKILL.md가 200줄 이하"
conflicts with the actual SKILL.md length (239 lines); update the criterion to
reflect the real target (e.g., change to 240 or 239 lines) or change the
guidance to require "≤240 lines" or "simplify SKILL.md to ≤200 lines" depending
on the intended policy, and make the same wording change wherever the exact
200-line threshold for SKILL.md is referenced in that spec so the document and
implementation are consistent.
---
Outside diff comments:
In `@README.md`:
- Line 49: Update the README so the "Personal skills (16 total)" entry for the
me plugin matches the actual skills: either (A) update the count to the real
number of skill folders found under plugins/me/skills/ (22 excluding setup) and
ensure the skills table lists all 22 names (including autoresearch), or (B)
prune the table to exactly match 16 skills if you intend to advertise 16 — make
the README table, the numeric total, and the actual plugins/me/skills/ directory
contents consistent; modify the skills table in README.md and the total label so
all three sources (count, table, directory) agree.
---
Nitpick comments:
In `@docs/superpowers/plans/2026-05-26-autoresearch-me-integration.md`:
- Around line 477-489: Add blank lines before and after the fenced code block
and include a language identifier on the opening fence; specifically, ensure
there is an empty line above the triple-backtick fence that contains the YAML
starting with "name: autoresearch" and an empty line after the closing fence,
and change the opening fence to something like "```yaml" so the block is
properly recognized as YAML.
In `@docs/superpowers/specs/2026-05-26-autoresearch-me-integration-design.md`:
- Around line 144-158: The fenced example block lacks a language identifier;
update the code fence that opens before "User: /me:autoresearch ..." to use a
language tag (e.g., ```text) so the snippet is rendered correctly; locate the
fenced block containing the "User: /me:autoresearch \"optimize parser perf, R² >
0.85\"" example and change the opening triple backticks to "```text" (and keep
the closing triple backticks) to follow the proposed fix.
- Around line 35-40: The fenced directory-structure code block in the docs
containing "plugins/me/skills/autoresearch/" lacks a language identifier; update
the opening fence to include a language (e.g., change ``` to ```text) so the
block is rendered without erroneous syntax highlighting—edit the block that
contains the lines "plugins/me/skills/autoresearch/", "└── SKILL.md", and
"plugins/autoresearch/ # 삭제" to use ```text as the fence.
- Around line 62-111: The top-level fenced code block that contains the
autoresearch spec currently lacks a language identifier; update the opening
fence (the ``` that precedes the line "# autoresearch: Autonomous experiment
loop") to include a language tag such as text or markdown (e.g., change ``` to
```text) so markdown renderers will format it correctly and preserve explicit
plaintext styling for the whole block.
🪄 Autofix (Beta)
Fix all unresolved CodeRabbit comments on this PR:
- Push a commit to this branch (recommended)
- Create a new PR with the fixes
ℹ️ Review info
⚙️ Run configuration
Configuration used: defaults
Review profile: CHILL
Plan: Pro
Run ID: 73fbdbea-165d-4bc3-91e0-ebbdf0dfc740
📒 Files selected for processing (17)
.agents/plugins/marketplace.json.claude-plugin/marketplace.json.github/workflows/sync-marketplace.ymlREADME.mddocs/superpowers/plans/2026-05-26-autoresearch-me-integration.mddocs/superpowers/specs/2026-05-26-autoresearch-me-integration-design.mdplugins/autoresearch/.claude-plugin/plugin.jsonplugins/autoresearch/.codex-plugin/plugin.jsonplugins/autoresearch/commands/autoresearch.mdplugins/autoresearch/hooks/autoresearch-context.shplugins/autoresearch/hooks/hooks.jsonplugins/autoresearch/skills/autoresearch/SKILL.mdplugins/me/skills/autoresearch/SKILL.mdplugins/me/skills/setup/settings.jsonscripts/check-codex-artifacts.shtests/codex_marketplace_json.batstests/codex_plugin_json.bats
💤 Files with no reviewable changes (11)
- plugins/autoresearch/hooks/autoresearch-context.sh
- plugins/autoresearch/hooks/hooks.json
- plugins/autoresearch/.codex-plugin/plugin.json
- plugins/me/skills/setup/settings.json
- .agents/plugins/marketplace.json
- plugins/autoresearch/skills/autoresearch/SKILL.md
- plugins/autoresearch/commands/autoresearch.md
- plugins/autoresearch/.claude-plugin/plugin.json
- scripts/check-codex-artifacts.sh
- .github/workflows/sync-marketplace.yml
- .claude-plugin/marketplace.json
| - [ ] 두 marketplace.json에 autoresearch 엔트리 없음 | ||
| - [ ] 모든 BATS 테스트 통과 | ||
| - [ ] pre-commit 통과 | ||
| - [ ] SKILL.md가 200줄 이하 (원본 254줄 대비 단순화 입증) |
There was a problem hiding this comment.
Success criterion conflicts with actual implementation.
The completion criterion states SKILL.md should be ≤200 lines, but the actual plugins/me/skills/autoresearch/SKILL.md being added is 239 lines. Either update this criterion to reflect the actual line count (e.g., "≤240 lines") or simplify the SKILL.md implementation to meet the 200-line target.
🔧 Proposed fix
-- [ ] SKILL.md가 200줄 이하 (원본 254줄 대비 단순화 입증)
+- [ ] SKILL.md가 240줄 이하 (원본 254줄 대비 단순화 입증)📝 Committable suggestion
‼️ IMPORTANT
Carefully review the code before committing. Ensure that it accurately replaces the highlighted code, contains no missing lines, and has no issues with indentation. Thoroughly test & benchmark the code to ensure it meets the requirements.
Suggested change
| - [ ] SKILL.md가 200줄 이하 (원본 254줄 대비 단순화 입증) | |
| - [ ] SKILL.md가 240줄 이하 (원본 254줄 대비 단순화 입증) |
🤖 Prompt for AI Agents
Verify each finding against current code. Fix only still-valid issues, skip the
rest with a brief reason, keep changes minimal, and validate.
In `@docs/superpowers/plans/2026-05-26-autoresearch-me-integration.md` at line
516, The checklist entry "[ ] SKILL.md가 200줄 이하" conflicts with the actual
SKILL.md (239 lines); either update the criterion text to match the real file
(e.g., change the checkbox label to "SKILL.md가 240줄 이하" or "≤240 lines") or
reduce the SKILL.md content to <=200 lines; locate the checklist line containing
"[ ] SKILL.md가 200줄 이하" and either edit that label to the new limit or refactor
the file named SKILL.md to remove/simplify sections until its line count is
<=200, ensuring the checkbox and the actual file length are consistent.
| ## 성공 기준 | ||
|
|
||
| - `plugins/autoresearch/` 디렉토리 제거되어도 `/autoresearch <goal>`이 me plugin을 통해 작동 | ||
| - SKILL.md가 200줄 이하 (현재 254줄) |
There was a problem hiding this comment.
Success criterion conflicts with actual implementation.
The criterion states SKILL.md should be ≤200 lines (down from 254), but the actual implementation is 239 lines. Update this criterion to match reality or simplify the SKILL.md.
🔧 Proposed fix
-- SKILL.md가 200줄 이하 (현재 254줄)
+- SKILL.md가 240줄 이하 (현재 254줄 대비 6% 감소)📝 Committable suggestion
‼️ IMPORTANT
Carefully review the code before committing. Ensure that it accurately replaces the highlighted code, contains no missing lines, and has no issues with indentation. Thoroughly test & benchmark the code to ensure it meets the requirements.
Suggested change
| - SKILL.md가 200줄 이하 (현재 254줄) | |
| - SKILL.md가 240줄 이하 (현재 254줄 대비 6% 감소) |
🤖 Prompt for AI Agents
Verify each finding against current code. Fix only still-valid issues, skip the
rest with a brief reason, keep changes minimal, and validate.
In `@docs/superpowers/specs/2026-05-26-autoresearch-me-integration-design.md` at
line 178, The success criterion in
2026-05-26-autoresearch-me-integration-design.md that states "SKILL.md가 200줄 이하"
conflicts with the actual SKILL.md length (239 lines); update the criterion to
reflect the real target (e.g., change to 240 or 239 lines) or change the
guidance to require "≤240 lines" or "simplify SKILL.md to ≤200 lines" depending
on the intended policy, and make the same wording change wherever the exact
200-line threshold for SKILL.md is referenced in that spec so the document and
implementation are consistent.
5 tasks
baleen37
added a commit
that referenced
this pull request
May 31, 2026
…#668) * docs(autoresearch): spec for restoring standalone plugin /goal is a user-only session command that the agent cannot execute, so the me-integrated skill breaks at the /goal step (Exit 127). Restore the pre-#660 standalone plugins/autoresearch/ which uses a UserPromptSubmit hook instead of /goal. * docs(autoresearch): implementation plan for standalone plugin restore * feat(autoresearch): restore standalone plugin files from efa9cdf~1 Extract the pre-#660 plugin (commands, UserPromptSubmit hook, skill) which uses a hook instead of the unrunnable /goal command. * chore(autoresearch): align plugin version to 17.16.2 * chore(me): remove autoresearch skill (moved to standalone plugin) * chore(marketplace): register autoresearch plugin * docs(readme): move autoresearch to its own plugin section * fix(marketplace): sync autoresearch into codex marketplace The codex marketplace at .agents/plugins/marketplace.json must mirror the claude marketplace skill plugins in order. Add autoresearch and bump the expected plugin count in the bats test from 4 to 5. * docs(autoresearch): record codex marketplace sync in plan --------- Co-authored-by: Test User <test@example.com>
baleen-release-app Bot
pushed a commit
that referenced
this pull request
May 31, 2026
# [17.17.0](v17.16.2...v17.17.0) (2026-05-31) ### Features * **autoresearch:** restore standalone plugin (drop unrunnable /goal) ([#668](#668)) ([7e92589](7e92589)), closes [pre-#660](https://github.com/pre-/issues/660) [pre-#660](https://github.com/pre-/issues/660)
This was referenced Jul 31, 2026
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
plugins/autoresearch/독립 plugin을plugins/me/skills/autoresearch/SKILL.md단일 파일로 통합/goal명령어 (v2.1.139+) 가 매 턴 종료 조건 평가 담당UserPromptSubmithook 폐기 (/goal이 턴 사이 자동 진행)Changes
plugins/me/skills/autoresearch/SKILL.md— 238 lines,/goal위임plugins/autoresearch/(6 파일).github/workflows/sync-marketplace.yml,scripts/check-codex-artifacts.sh, 2개 codex BATS 테스트,plugins/me/skills/setup/settings.json,README.mdBackground
조사 결과 obra/autoresearch upstream은 2026-03-18 이후 변경 없음. 가져올 변경사항 없음. 대신 Claude Code
/goal이 현재 plugin의 핵심 메커니즘(턴 사이 자동 진행)을 깔끔하게 대체.Spec:
docs/superpowers/specs/2026-05-26-autoresearch-me-integration-design.mdPlan:
docs/superpowers/plans/2026-05-26-autoresearch-me-integration.mdTest plan
bats tests/— 71/71 passingjq emptyon both marketplace.json filesplugins/autoresearchreferences (grep clean)plugins/me/skills/autoresearch/SKILL.mdfrontmatter 유효/me:autoresearch <goal>호출 시/goal자동 설정 (consumer 환경에서)Notes
/goal명령어). SKILL.md Prerequisites 섹션에 명기.Summary by CodeRabbit
Refactor
Documentation
Tests