[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"skill:01M7FDD37N4XDBR1PTSZCKV0TP":3,"skill-same:01M7FDD37N4XDBR1PTSZCKV0TP":46},{"skillNo":4,"name":5,"nameZh":6,"description":7,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":16,"categories":17,"keywords":18,"publishedAt":19,"linkOnly":20,"markdownContent":21,"markdownBody":22,"frontMatter":23,"refName":24,"commitSha":25,"skillDirectory":26,"forksCount":27,"repositoryHtmlUrl":28,"repositoryHomepageUrl":6,"ownerHtmlUrl":29,"ownerBlogUrl":6,"ownerDescription":30,"updatedAt":31,"install":32,"resources":45},"01M7FDD37N4XDBR1PTSZCKV0TP","auto-review-loop",null,"Autonomous multi-round research review loop. Repeatedly reviews using Gemini via gemini-review MCP, implements fixes, and re-reviews until positive assessment or max rounds reached. Use when user says \"auto review loop\", \"review until it passes\", or wants autonomous iterative improvement.","CZX-LBJ\u002Fdeepseek-science","CZX-LBJ","https:\u002F\u002Favatars.githubusercontent.com\u002Fu\u002F204594689?v=4","01M7FC59WQ9PFXT04MKGJDCFFT","czx","01M7FC59XB2GW0BKWN9JTDBQS7",1,"MIT","https:\u002F\u002Fgithub.com\u002FCZX-LBJ\u002Fdeepseek-science\u002Fblob\u002Fmain\u002Fpackages\u002Fskill\u002Fskill-science\u002Fassets\u002Fsources\u002Faris\u002Fskills\u002Fskills-codex-gemini-review\u002Fauto-review-loop\u002FSKILL.md",[],[],"2026-11-16 10:05:34",false,"---\nname: \"auto-review-loop\"\ndescription: \"Autonomous multi-round research review loop. Repeatedly reviews using Gemini via gemini-review MCP, implements fixes, and re-reviews until positive assessment or max rounds reached. Use when user says \\\"auto review loop\\\", \\\"review until it passes\\\", or wants autonomous iterative improvement.\"\n---\n\n> Override for Codex users who want **Gemini**, not a second Codex agent, to act as the reviewer. Install this package **after** `skills\u002Fskills-codex\u002F*`.\n\n# Auto Review Loop: Autonomous Research Improvement\n\n> **Gemini overlay assurance:** `review_independence: cross-family` and `acceptance_status: accepted`.\n\nAutonomously iterate: review → implement fixes → re-review, until the external reviewer gives a positive assessment or MAX_ROUNDS is reached.\n\n## Context: $ARGUMENTS\n\n## Constants\n\n- MAX_ROUNDS = 4\n- POSITIVE_THRESHOLD: score >= 6\u002F10 AND verdict ∈ {\"ready\", \"almost\"} — both must hold, matching the operative STOP CONDITION below. Verdict vocabulary is {\"ready\", \"almost\", \"not ready\"}. (Earlier wording used \"or\" + a stale verdict set; the AND form is authoritative.)\n- REVIEW_DOC: `review-stage\u002FAUTO_REVIEW.md` (cumulative log) *(fall back to `.\u002FAUTO_REVIEW.md` for legacy projects)*\n- **OUTPUT_DIR = `review-stage\u002F`** — Directory for review output files.\n- **REVIEWER_MODEL = `gemini-review`** — Gemini reviewer invoked through the local `gemini-review` MCP bridge. Set `GEMINI_REVIEW_MODEL` if you need a specific Gemini model override.\n- **HUMAN_CHECKPOINT = false** — When `true`, pause after each round's review (Phase B) and present the score + weaknesses to the user. Wait for user input before proceeding to Phase C. The user can: approve the suggested fixes, provide custom modification instructions, skip specific fixes, or stop the loop early. When `false` (default), the loop runs fully autonomously.\n\n> 💡 Override: `\u002Fauto-review-loop \"topic\" — human checkpoint: true`\n\n## State Persistence (Compact Recovery)\n\nLong-running loops may hit the context window limit, triggering automatic compaction. To survive this, persist state to `review-stage\u002FREVIEW_STATE.json` after each round:\n\n```json\n{\n  \"round\": 2,\n  \"thread_id\": \"019cd392-...\",\n  \"status\": \"in_progress\",\n  \"last_score\": 5.0,\n  \"last_verdict\": \"not ready\",\n  \"pending_experiments\": [\"screen_name_1\"],\n  \"timestamp\": \"2026-03-13T21:00:00\"\n}\n```\n\n**Write this file at the end of every Phase E** (after documenting the round). Overwrite each time — only the latest state matters.\n\n**On completion** (positive assessment or max rounds), set `\"status\": \"completed\"` so future invocations don't accidentally resume a finished loop.\n\n## Workflow\n\n### Initialization\n\n1. **Check for `review-stage\u002FREVIEW_STATE.json`** *(fall back to `.\u002FREVIEW_STATE.json` if not found — legacy path)*:\n   - If neither path exists: **fresh start** (normal case, identical to behavior before this feature existed)\n   - If it exists AND `status` is `\"completed\"`: **fresh start** (previous loop finished normally)\n   - If it exists AND `status` is `\"in_progress\"` AND `timestamp` is older than 24 hours: **fresh start** (stale state from a killed\u002Fabandoned run — delete the file and start over)\n   - If it exists AND `status` is `\"in_progress\"` AND `timestamp` is within 24 hours: **resume**\n     - Read the state file to recover `round`, `thread_id`, `last_score`, `pending_experiments`\n     - Read `review-stage\u002FAUTO_REVIEW.md` to restore full context of prior rounds *(fall back to `.\u002FAUTO_REVIEW.md`)*\n     - If `pending_experiments` is non-empty, check if they have completed (e.g., check screen sessions)\n     - Resume from the next round (round = saved round + 1)\n     - Log: \"Recovered from context compaction. Resuming at Round N.\"\n2. Read project narrative documents, memory files, and any prior review documents\n3. Read recent experiment results (check output directories, logs)\n4. Identify current weaknesses and open TODOs from prior reviews\n5. Initialize round counter = 1 (unless recovered from state file)\n6. Create\u002Fupdate `review-stage\u002FAUTO_REVIEW.md` with header and timestamp\n\n### Loop (repeat up to MAX_ROUNDS)\n\n#### Phase A: Review\n\nSend comprehensive context to the external reviewer:\n\n```\nmcp__gemini-review__review_start:\n  prompt: |\n    [Round N\u002FMAX_ROUNDS of autonomous review loop]\n\n    [Full research context: claims, methods, results, known weaknesses]\n    [Changes since last round, if any]\n\n    Please act as a senior ML reviewer (NeurIPS\u002FICML level).\n\n    1. Score this work 1-10 for a top venue\n    2. List remaining critical weaknesses (ranked by severity)\n    3. For each weakness, specify the MINIMUM fix (experiment, analysis, or reframing)\n    4. State clearly: is this READY for submission? Yes\u002FNo\u002FAlmost\n\n    Be brutally honest. If the work is ready, say so clearly.\n\n    === SCOPE LIMITS (these bound what you PROPOSE, never what you look for) ===\n    Report anything that is actually wrong here — including a rare-looking case, if\n    this repo actually produces it. Then keep the fix in scope:\n    1. This is a RESEARCH-WORKFLOW tool, not a security paper. Verification is\n       welcome; over-defense is not. Assume a cooperating operator on their own\n       machine — a malicious local user is NOT in the threat model.\n    2. Do NOT propose SHA \u002F hash \u002F content-fingerprint \u002F digest-binding schemes.\n       Reporting a real defect in hashing code that already exists is fine.\n    3. NO speculative machinery: do not add feature flags, migration frameworks,\n       compat layers, wrappers, pins, or similar mechanisms unless evidence shows\n       a current repo defect they fix or an explicit existing invariant they must\n       preserve. \"Load-bearing\", \"compatibility\", and \"not scaffolding\" are labels,\n       not evidence. Point to the failing path\u002Fartifact or invariant, and check the\n       proposal's factual premises, such as whether a named package version exists.\n    4. NO corner-case obsession: exotic encodings, symlink races, RTL text and\n       millisecond races are out of scope unless you can show the case arises here.\n    5. Where a rubric or checklist is genuinely needed, do not over-mechanize\n       judgement. A clear sentence a human reads beats a scored table nobody\n       maintains.\n    Exception: code that runs remote commands, starts a network service, or installs\n    an MCP server runs on the user's machine with their credentials — trust-boundary\n    findings there are in scope and the default is strict.\n    Say plainly when something is correct. Do not manufacture findings.\n```\n\nAfter this start call, immediately save the returned `jobId` and poll `mcp__gemini-review__review_status` with a bounded `waitSeconds` until `done=true`. Treat the completed status payload's `response` as the reviewer output, and save the completed `threadId` for any follow-up round.\n\nIf this is round 2+, use `mcp__gemini-review__review_reply_start` with the saved completed `threadId`, then poll `mcp__gemini-review__review_status` with the returned `jobId` until `done=true` to maintain continuity.\n\n#### Phase B: Parse Assessment\n\n**CRITICAL: Save the FULL raw response** from the external reviewer verbatim (store in a variable for Phase E). Do NOT discard or summarize — the raw text is the primary record.\n\nThen extract structured fields:\n- **Score** (numeric 1-10)\n- **Verdict** (\"ready\" \u002F \"almost\" \u002F \"not ready\")\n- **Action items** (ranked list of fixes)\n\n**STOP CONDITION**: If score >= 6 AND verdict ∈ {\"ready\", \"almost\"} (exact match — \"not ready\" does NOT qualify) → stop loop, document final state.\n\n#### Human Checkpoint (if enabled)\n\n**Skip this step entirely if `HUMAN_CHECKPOINT = false`.**\n\nWhen `HUMAN_CHECKPOINT = true`, present the review results and wait for user input:\n\n```\n📋 Round N\u002FMAX_ROUNDS review complete.\n\nScore: X\u002F10 — [verdict]\nTop weaknesses:\n1. [weakness 1]\n2. [weakness 2]\n3. [weakness 3]\n\nSuggested fixes:\n1. [fix 1]\n2. [fix 2]\n3. [fix 3]\n\nOptions:\n- Reply \"go\" or \"continue\" → implement all suggested fixes\n- Reply with custom instructions → implement your modifications instead\n- Reply \"skip 2\" → skip fix #2, implement the rest\n- Reply \"stop\" → end the loop, document current state\n```\n\nWait for the user's response. Parse their input:\n- **Approval** (\"go\", \"continue\", \"ok\", \"proceed\"): proceed to Phase C with all suggested fixes\n- **Custom instructions** (any other text): treat as additional\u002Freplacement guidance for Phase C. Merge with reviewer suggestions where appropriate\n- **Skip specific fixes** (\"skip 1,3\"): remove those fixes from the action list\n- **Stop** (\"stop\", \"enough\", \"done\"): terminate the loop, jump to Termination\n\n#### Feishu Notification (if configured)\n\nAfter parsing the score, check if `~\u002F.codex\u002Ffeishu.json` exists and mode is not `\"off\"`:\n- Send a `review_scored` notification: \"Round N: X\u002F10 — [verdict]\" with top 3 weaknesses\n- If **interactive** mode and verdict is \"almost\": send as checkpoint, wait for user reply on whether to continue or stop\n- If config absent or mode off: skip entirely (no-op)\n\n#### Phase C: Implement Fixes (if not stopping)\n\nFor each action item (highest priority first):\n\n1. **Code changes**: Write\u002Fmodify experiment scripts, model code, analysis scripts\n2. **Run experiments**: Deploy to GPU server via SSH + screen\u002Ftmux\n3. **Analysis**: Run evaluation, collect results, update figures\u002Ftables\n4. **Documentation**: Update project notes and review document\n\nPrioritization rules:\n- Skip fixes requiring excessive compute (flag for manual follow-up)\n- Skip fixes requiring external data\u002Fmodels not available\n- Prefer reframing\u002Fanalysis over new experiments when both address the concern\n- Always implement metric additions (cheap, high impact)\n\n#### Phase D: Wait for Results\n\nIf experiments were launched:\n- Monitor remote sessions for completion\n- Collect results from output files and logs\n\n#### Phase E: Document Round\n\nAppend to `review-stage\u002FAUTO_REVIEW.md`:\n\n```markdown\n## Round N (timestamp)\n\n### Assessment (Summary)\n- Score: X\u002F10\n- Verdict: [ready\u002Falmost\u002Fnot ready]\n- Key criticisms: [bullet list]\n\n### Reviewer Raw Response\n\n\u003Cdetails>\n\u003Csummary>Click to expand full reviewer response\u003C\u002Fsummary>\n\n[Paste the COMPLETE raw response from the external reviewer here — verbatim, unedited.\nThis is the authoritative record. Do NOT truncate or paraphrase.]\n\n\u003C\u002Fdetails>\n\n### Actions Taken\n- [what was implemented\u002Fchanged]\n\n### Results\n- [experiment outcomes, if any]\n\n### Status\n- [continuing to round N+1 \u002F stopping]\n```\n\n**Write `review-stage\u002FREVIEW_STATE.json`** with current round, agent id, score, verdict, and any pending experiments.\n\nIncrement round counter → back to Phase A.\n\n### Termination\n\nWhen loop ends (positive assessment or max rounds):\n\n1. Update `review-stage\u002FREVIEW_STATE.json` with `\"status\": \"completed\"`\n2. Write final summary to `review-stage\u002FAUTO_REVIEW.md`\n3. Update project notes with conclusions\n4. If stopped at max rounds without positive assessment:\n   - List remaining blockers\n   - Estimate effort needed for each\n   - Suggest whether to continue manually or pivot\n5. **Feishu notification** (if configured): Send `pipeline_done` with final score progression table\n\n## Output Protocols\n\n> Follow these shared protocols for all output files:\n> - **[Output Versioning Protocol](..\u002F..\u002Fshared-references\u002Foutput-versioning.md)** — write timestamped file first, then copy to fixed name\n> - **[Output Manifest Protocol](..\u002F..\u002Fshared-references\u002Foutput-manifest.md)** — log every output to MANIFEST.md\n> - **[Output Language Protocol](..\u002F..\u002Fshared-references\u002Foutput-language.md)** — respect the project's language setting\n\n## Key Rules\n\n- **Large file handling**: If the Write tool fails due to file size, immediately retry using Bash (`cat \u003C\u003C 'EOF' > file`) to write in chunks. Do NOT ask the user for permission — just do it silently.\n\n- Always ask the Gemini reviewer for strict, high-rigor feedback.\n- Save the completed `threadId` from the first `mcp__gemini-review__review_status` result, then use `mcp__gemini-review__review_reply_start` plus `mcp__gemini-review__review_status` for subsequent rounds\n- Be honest — include negative results and failed experiments\n- Do NOT hide weaknesses to game a positive score\n- Implement fixes BEFORE re-reviewing (don't just promise to fix)\n- If an experiment takes > 30 minutes, launch it and continue with other fixes while waiting\n- Document EVERYTHING — the review log should be self-contained\n- Update project notes after each round, not just at the end\n\n## Prompt Template for Round 2+\n\n```\nmcp__gemini-review__review_reply_start:\n  threadId: [saved from round 1]\n  prompt: |\n    [Round N update]\n\n    Since your last review, we have:\n    1. [Action 1]: [result]\n    2. [Action 2]: [result]\n    3. [Action 3]: [result]\n\n    Updated results table:\n    [paste metrics]\n\n    Please re-score and re-assess. Are the remaining concerns addressed?\n    Same format: Score, Verdict, Remaining Weaknesses, Minimum Fixes.\n\n    === SCOPE LIMITS (these bound what you PROPOSE, never what you look for) ===\n    Report anything that is actually wrong here — including a rare-looking case, if\n    this repo actually produces it. Then keep the fix in scope:\n    1. This is a RESEARCH-WORKFLOW tool, not a security paper. Verification is\n       welcome; over-defense is not. Assume a cooperating operator on their own\n       machine — a malicious local user is NOT in the threat model.\n    2. Do NOT propose SHA \u002F hash \u002F content-fingerprint \u002F digest-binding schemes.\n       Reporting a real defect in hashing code that already exists is fine.\n    3. NO speculative machinery: do not add feature flags, migration frameworks,\n       compat layers, wrappers, pins, or similar mechanisms unless evidence shows\n       a current repo defect they fix or an explicit existing invariant they must\n       preserve. \"Load-bearing\", \"compatibility\", and \"not scaffolding\" are labels,\n       not evidence. Point to the failing path\u002Fartifact or invariant, and check the\n       proposal's factual premises, such as whether a named package version exists.\n    4. NO corner-case obsession: exotic encodings, symlink races, RTL text and\n       millisecond races are out of scope unless you can show the case arises here.\n    5. Where a rubric or checklist is genuinely needed, do not over-mechanize\n       judgement. A clear sentence a human reads beats a scored table nobody\n       maintains.\n    Exception: code that runs remote commands, starts a network service, or installs\n    an MCP server runs on the user's machine with their credentials — trust-boundary\n    findings there are in scope and the default is strict.\n    Say plainly when something is correct. Do not manufacture findings.\n```\n\nAfter this start call, immediately save the returned `jobId` and poll `mcp__gemini-review__review_status` with a bounded `waitSeconds` until `done=true`. Treat the completed status payload's `response` as the reviewer output, and save the completed `threadId` for any follow-up round.\n","> Override for Codex users who want **Gemini**, not a second Codex agent, to act as the reviewer. Install this package **after** `skills\u002Fskills-codex\u002F*`.\n\n# Auto Review Loop: Autonomous Research Improvement\n\n> **Gemini overlay assurance:** `review_independence: cross-family` and `acceptance_status: accepted`.\n\nAutonomously iterate: review → implement fixes → re-review, until the external reviewer gives a positive assessment or MAX_ROUNDS is reached.\n\n## Context: $ARGUMENTS\n\n## Constants\n\n- MAX_ROUNDS = 4\n- POSITIVE_THRESHOLD: score >= 6\u002F10 AND verdict ∈ {\"ready\", \"almost\"} — both must hold, matching the operative STOP CONDITION below. Verdict vocabulary is {\"ready\", \"almost\", \"not ready\"}. (Earlier wording used \"or\" + a stale verdict set; the AND form is authoritative.)\n- REVIEW_DOC: `review-stage\u002FAUTO_REVIEW.md` (cumulative log) *(fall back to `.\u002FAUTO_REVIEW.md` for legacy projects)*\n- **OUTPUT_DIR = `review-stage\u002F`** — Directory for review output files.\n- **REVIEWER_MODEL = `gemini-review`** — Gemini reviewer invoked through the local `gemini-review` MCP bridge. Set `GEMINI_REVIEW_MODEL` if you need a specific Gemini model override.\n- **HUMAN_CHECKPOINT = false** — When `true`, pause after each round's review (Phase B) and present the score + weaknesses to the user. Wait for user input before proceeding to Phase C. The user can: approve the suggested fixes, provide custom modification instructions, skip specific fixes, or stop the loop early. When `false` (default), the loop runs fully autonomously.\n\n> 💡 Override: `\u002Fauto-review-loop \"topic\" — human checkpoint: true`\n\n## State Persistence (Compact Recovery)\n\nLong-running loops may hit the context window limit, triggering automatic compaction. To survive this, persist state to `review-stage\u002FREVIEW_STATE.json` after each round:\n\n```json\n{\n  \"round\": 2,\n  \"thread_id\": \"019cd392-...\",\n  \"status\": \"in_progress\",\n  \"last_score\": 5.0,\n  \"last_verdict\": \"not ready\",\n  \"pending_experiments\": [\"screen_name_1\"],\n  \"timestamp\": \"2026-03-13T21:00:00\"\n}\n```\n\n**Write this file at the end of every Phase E** (after documenting the round). Overwrite each time — only the latest state matters.\n\n**On completion** (positive assessment or max rounds), set `\"status\": \"completed\"` so future invocations don't accidentally resume a finished loop.\n\n## Workflow\n\n### Initialization\n\n1. **Check for `review-stage\u002FREVIEW_STATE.json`** *(fall back to `.\u002FREVIEW_STATE.json` if not found — legacy path)*:\n   - If neither path exists: **fresh start** (normal case, identical to behavior before this feature existed)\n   - If it exists AND `status` is `\"completed\"`: **fresh start** (previous loop finished normally)\n   - If it exists AND `status` is `\"in_progress\"` AND `timestamp` is older than 24 hours: **fresh start** (stale state from a killed\u002Fabandoned run — delete the file and start over)\n   - If it exists AND `status` is `\"in_progress\"` AND `timestamp` is within 24 hours: **resume**\n     - Read the state file to recover `round`, `thread_id`, `last_score`, `pending_experiments`\n     - Read `review-stage\u002FAUTO_REVIEW.md` to restore full context of prior rounds *(fall back to `.\u002FAUTO_REVIEW.md`)*\n     - If `pending_experiments` is non-empty, check if they have completed (e.g., check screen sessions)\n     - Resume from the next round (round = saved round + 1)\n     - Log: \"Recovered from context compaction. Resuming at Round N.\"\n2. Read project narrative documents, memory files, and any prior review documents\n3. Read recent experiment results (check output directories, logs)\n4. Identify current weaknesses and open TODOs from prior reviews\n5. Initialize round counter = 1 (unless recovered from state file)\n6. Create\u002Fupdate `review-stage\u002FAUTO_REVIEW.md` with header and timestamp\n\n### Loop (repeat up to MAX_ROUNDS)\n\n#### Phase A: Review\n\nSend comprehensive context to the external reviewer:\n\n```\nmcp__gemini-review__review_start:\n  prompt: |\n    [Round N\u002FMAX_ROUNDS of autonomous review loop]\n\n    [Full research context: claims, methods, results, known weaknesses]\n    [Changes since last round, if any]\n\n    Please act as a senior ML reviewer (NeurIPS\u002FICML level).\n\n    1. Score this work 1-10 for a top venue\n    2. List remaining critical weaknesses (ranked by severity)\n    3. For each weakness, specify the MINIMUM fix (experiment, analysis, or reframing)\n    4. State clearly: is this READY for submission? Yes\u002FNo\u002FAlmost\n\n    Be brutally honest. If the work is ready, say so clearly.\n\n    === SCOPE LIMITS (these bound what you PROPOSE, never what you look for) ===\n    Report anything that is actually wrong here — including a rare-looking case, if\n    this repo actually produces it. Then keep the fix in scope:\n    1. This is a RESEARCH-WORKFLOW tool, not a security paper. Verification is\n       welcome; over-defense is not. Assume a cooperating operator on their own\n       machine — a malicious local user is NOT in the threat model.\n    2. Do NOT propose SHA \u002F hash \u002F content-fingerprint \u002F digest-binding schemes.\n       Reporting a real defect in hashing code that already exists is fine.\n    3. NO speculative machinery: do not add feature flags, migration frameworks,\n       compat layers, wrappers, pins, or similar mechanisms unless evidence shows\n       a current repo defect they fix or an explicit existing invariant they must\n       preserve. \"Load-bearing\", \"compatibility\", and \"not scaffolding\" are labels,\n       not evidence. Point to the failing path\u002Fartifact or invariant, and check the\n       proposal's factual premises, such as whether a named package version exists.\n    4. NO corner-case obsession: exotic encodings, symlink races, RTL text and\n       millisecond races are out of scope unless you can show the case arises here.\n    5. Where a rubric or checklist is genuinely needed, do not over-mechanize\n       judgement. A clear sentence a human reads beats a scored table nobody\n       maintains.\n    Exception: code that runs remote commands, starts a network service, or installs\n    an MCP server runs on the user's machine with their credentials — trust-boundary\n    findings there are in scope and the default is strict.\n    Say plainly when something is correct. Do not manufacture findings.\n```\n\nAfter this start call, immediately save the returned `jobId` and poll `mcp__gemini-review__review_status` with a bounded `waitSeconds` until `done=true`. Treat the completed status payload's `response` as the reviewer output, and save the completed `threadId` for any follow-up round.\n\nIf this is round 2+, use `mcp__gemini-review__review_reply_start` with the saved completed `threadId`, then poll `mcp__gemini-review__review_status` with the returned `jobId` until `done=true` to maintain continuity.\n\n#### Phase B: Parse Assessment\n\n**CRITICAL: Save the FULL raw response** from the external reviewer verbatim (store in a variable for Phase E). Do NOT discard or summarize — the raw text is the primary record.\n\nThen extract structured fields:\n- **Score** (numeric 1-10)\n- **Verdict** (\"ready\" \u002F \"almost\" \u002F \"not ready\")\n- **Action items** (ranked list of fixes)\n\n**STOP CONDITION**: If score >= 6 AND verdict ∈ {\"ready\", \"almost\"} (exact match — \"not ready\" does NOT qualify) → stop loop, document final state.\n\n#### Human Checkpoint (if enabled)\n\n**Skip this step entirely if `HUMAN_CHECKPOINT = false`.**\n\nWhen `HUMAN_CHECKPOINT = true`, present the review results and wait for user input:\n\n```\n📋 Round N\u002FMAX_ROUNDS review complete.\n\nScore: X\u002F10 — [verdict]\nTop weaknesses:\n1. [weakness 1]\n2. [weakness 2]\n3. [weakness 3]\n\nSuggested fixes:\n1. [fix 1]\n2. [fix 2]\n3. [fix 3]\n\nOptions:\n- Reply \"go\" or \"continue\" → implement all suggested fixes\n- Reply with custom instructions → implement your modifications instead\n- Reply \"skip 2\" → skip fix #2, implement the rest\n- Reply \"stop\" → end the loop, document current state\n```\n\nWait for the user's response. Parse their input:\n- **Approval** (\"go\", \"continue\", \"ok\", \"proceed\"): proceed to Phase C with all suggested fixes\n- **Custom instructions** (any other text): treat as additional\u002Freplacement guidance for Phase C. Merge with reviewer suggestions where appropriate\n- **Skip specific fixes** (\"skip 1,3\"): remove those fixes from the action list\n- **Stop** (\"stop\", \"enough\", \"done\"): terminate the loop, jump to Termination\n\n#### Feishu Notification (if configured)\n\nAfter parsing the score, check if `~\u002F.codex\u002Ffeishu.json` exists and mode is not `\"off\"`:\n- Send a `review_scored` notification: \"Round N: X\u002F10 — [verdict]\" with top 3 weaknesses\n- If **interactive** mode and verdict is \"almost\": send as checkpoint, wait for user reply on whether to continue or stop\n- If config absent or mode off: skip entirely (no-op)\n\n#### Phase C: Implement Fixes (if not stopping)\n\nFor each action item (highest priority first):\n\n1. **Code changes**: Write\u002Fmodify experiment scripts, model code, analysis scripts\n2. **Run experiments**: Deploy to GPU server via SSH + screen\u002Ftmux\n3. **Analysis**: Run evaluation, collect results, update figures\u002Ftables\n4. **Documentation**: Update project notes and review document\n\nPrioritization rules:\n- Skip fixes requiring excessive compute (flag for manual follow-up)\n- Skip fixes requiring external data\u002Fmodels not available\n- Prefer reframing\u002Fanalysis over new experiments when both address the concern\n- Always implement metric additions (cheap, high impact)\n\n#### Phase D: Wait for Results\n\nIf experiments were launched:\n- Monitor remote sessions for completion\n- Collect results from output files and logs\n\n#### Phase E: Document Round\n\nAppend to `review-stage\u002FAUTO_REVIEW.md`:\n\n```markdown\n## Round N (timestamp)\n\n### Assessment (Summary)\n- Score: X\u002F10\n- Verdict: [ready\u002Falmost\u002Fnot ready]\n- Key criticisms: [bullet list]\n\n### Reviewer Raw Response\n\n\u003Cdetails>\n\u003Csummary>Click to expand full reviewer response\u003C\u002Fsummary>\n\n[Paste the COMPLETE raw response from the external reviewer here — verbatim, unedited.\nThis is the authoritative record. Do NOT truncate or paraphrase.]\n\n\u003C\u002Fdetails>\n\n### Actions Taken\n- [what was implemented\u002Fchanged]\n\n### Results\n- [experiment outcomes, if any]\n\n### Status\n- [continuing to round N+1 \u002F stopping]\n```\n\n**Write `review-stage\u002FREVIEW_STATE.json`** with current round, agent id, score, verdict, and any pending experiments.\n\nIncrement round counter → back to Phase A.\n\n### Termination\n\nWhen loop ends (positive assessment or max rounds):\n\n1. Update `review-stage\u002FREVIEW_STATE.json` with `\"status\": \"completed\"`\n2. Write final summary to `review-stage\u002FAUTO_REVIEW.md`\n3. Update project notes with conclusions\n4. If stopped at max rounds without positive assessment:\n   - List remaining blockers\n   - Estimate effort needed for each\n   - Suggest whether to continue manually or pivot\n5. **Feishu notification** (if configured): Send `pipeline_done` with final score progression table\n\n## Output Protocols\n\n> Follow these shared protocols for all output files:\n> - **[Output Versioning Protocol](..\u002F..\u002Fshared-references\u002Foutput-versioning.md)** — write timestamped file first, then copy to fixed name\n> - **[Output Manifest Protocol](..\u002F..\u002Fshared-references\u002Foutput-manifest.md)** — log every output to MANIFEST.md\n> - **[Output Language Protocol](..\u002F..\u002Fshared-references\u002Foutput-language.md)** — respect the project's language setting\n\n## Key Rules\n\n- **Large file handling**: If the Write tool fails due to file size, immediately retry using Bash (`cat \u003C\u003C 'EOF' > file`) to write in chunks. Do NOT ask the user for permission — just do it silently.\n\n- Always ask the Gemini reviewer for strict, high-rigor feedback.\n- Save the completed `threadId` from the first `mcp__gemini-review__review_status` result, then use `mcp__gemini-review__review_reply_start` plus `mcp__gemini-review__review_status` for subsequent rounds\n- Be honest — include negative results and failed experiments\n- Do NOT hide weaknesses to game a positive score\n- Implement fixes BEFORE re-reviewing (don't just promise to fix)\n- If an experiment takes > 30 minutes, launch it and continue with other fixes while waiting\n- Document EVERYTHING — the review log should be self-contained\n- Update project notes after each round, not just at the end\n\n## Prompt Template for Round 2+\n\n```\nmcp__gemini-review__review_reply_start:\n  threadId: [saved from round 1]\n  prompt: |\n    [Round N update]\n\n    Since your last review, we have:\n    1. [Action 1]: [result]\n    2. [Action 2]: [result]\n    3. [Action 3]: [result]\n\n    Updated results table:\n    [paste metrics]\n\n    Please re-score and re-assess. Are the remaining concerns addressed?\n    Same format: Score, Verdict, Remaining Weaknesses, Minimum Fixes.\n\n    === SCOPE LIMITS (these bound what you PROPOSE, never what you look for) ===\n    Report anything that is actually wrong here — including a rare-looking case, if\n    this repo actually produces it. Then keep the fix in scope:\n    1. This is a RESEARCH-WORKFLOW tool, not a security paper. Verification is\n       welcome; over-defense is not. Assume a cooperating operator on their own\n       machine — a malicious local user is NOT in the threat model.\n    2. Do NOT propose SHA \u002F hash \u002F content-fingerprint \u002F digest-binding schemes.\n       Reporting a real defect in hashing code that already exists is fine.\n    3. NO speculative machinery: do not add feature flags, migration frameworks,\n       compat layers, wrappers, pins, or similar mechanisms unless evidence shows\n       a current repo defect they fix or an explicit existing invariant they must\n       preserve. \"Load-bearing\", \"compatibility\", and \"not scaffolding\" are labels,\n       not evidence. Point to the failing path\u002Fartifact or invariant, and check the\n       proposal's factual premises, such as whether a named package version exists.\n    4. NO corner-case obsession: exotic encodings, symlink races, RTL text and\n       millisecond races are out of scope unless you can show the case arises here.\n    5. Where a rubric or checklist is genuinely needed, do not over-mechanize\n       judgement. A clear sentence a human reads beats a scored table nobody\n       maintains.\n    Exception: code that runs remote commands, starts a network service, or installs\n    an MCP server runs on the user's machine with their credentials — trust-boundary\n    findings there are in scope and the default is strict.\n    Say plainly when something is correct. Do not manufacture findings.\n```\n\nAfter this start call, immediately save the returned `jobId` and poll `mcp__gemini-review__review_status` with a bounded `waitSeconds` until `done=true`. Treat the completed status payload's `response` as the reviewer output, and save the completed `threadId` for any follow-up round.\n",{"name":5,"description":7},"main","18193d71ea3441d7c033dbf195b8a7c75ac96f75","packages\u002Fskill\u002Fskill-science\u002Fassets\u002Fsources\u002Faris\u002Fskills\u002Fskills-codex-gemini-review\u002Fauto-review-loop",0,"https:\u002F\u002Fgithub.com\u002FCZX-LBJ\u002Fdeepseek-science","https:\u002F\u002Fgithub.com\u002FCZX-LBJ","一个没事喜欢随便写点代码的小菜鸡\r\n","2026-10-22 09:48:04",{"source":8,"skillName":5,"commands":33,"prompt":43,"zipAvailable":44},[34,37,40],{"tool":35,"command":36},"npx","npx skills add CZX-LBJ\u002Fdeepseek-science --skill auto-review-loop",{"tool":38,"command":39},"bunx","bunx skills add CZX-LBJ\u002Fdeepseek-science --skill auto-review-loop",{"tool":41,"command":42},"pnpm","pnpm dlx skills add CZX-LBJ\u002Fdeepseek-science --skill auto-review-loop","安装技能，技能地址：https:\u002F\u002Fgithub.com\u002FCZX-LBJ\u002Fdeepseek-science\u002Fblob\u002Fmain\u002Fpackages\u002Fskill\u002Fskill-science\u002Fassets\u002Fsources\u002Faris\u002Fskills\u002Fskills-codex-gemini-review\u002Fauto-review-loop\u002FSKILL.md",true,[],[47,54,61,68,75,82],{"skillNo":48,"name":49,"nameZh":6,"description":50,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":51,"categories":52,"keywords":53,"publishedAt":19},"01M7FEEKSNMRMKW6JEK93PFDFY","bio-free-energy-calculations","Performs alchemical free-energy calculations including relative binding free energy (RBFE \u002F FEP+) and absolute binding free energy (ABFE) via OpenFE, FEP+, GROMACS, AMBER pmemd, and OpenMM with explicit lambda scheduling, soft-core potentials, MBAR\u002FBAR analysis, cycle-closure validation, and protocol-appropriate enhanced sampling. Compares ML alternatives (Boltz-2 affinity, DeepDock). Use when ranking analogs by binding affinity beyond docking accuracy, performing prospective lead optimization, or validating SAR predictions.","https:\u002F\u002Fgithub.com\u002FCZX-LBJ\u002Fdeepseek-science\u002Fblob\u002Fmain\u002Fpackages\u002Fskill\u002Fskill-science\u002Fassets\u002Fsources\u002Fbioskills\u002Fchemoinformatics\u002Ffree-energy-calculations\u002FSKILL.md",[],[],{"skillNo":55,"name":56,"nameZh":6,"description":57,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":58,"categories":59,"keywords":60,"publishedAt":19},"01M7FEE5CTE5QK55CEMXNNVWY9","bio-conformer-generation","Generates 3D conformer ensembles using RDKit ETKDGv3 with knowledge-enhanced distance geometry, MMFF94\u002FUFF force-field optimization, CREST + GFN2-xTB semi-empirical refinement, and macrocycle-aware torsion preferences. Provides explicit decision rules for single vs ensemble conformer use, RMSD pruning, energy windows, conformer count, and force-field choice. Use when preparing 3D ligands for docking, generating descriptor input for 3D QSAR, or sampling macrocycle\u002Fpeptide conformational ensembles.","https:\u002F\u002Fgithub.com\u002FCZX-LBJ\u002Fdeepseek-science\u002Fblob\u002Fmain\u002Fpackages\u002Fskill\u002Fskill-science\u002Fassets\u002Fsources\u002Fbioskills\u002Fchemoinformatics\u002Fconformer-generation\u002FSKILL.md",[],[],{"skillNo":62,"name":63,"nameZh":6,"description":64,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":65,"categories":66,"keywords":67,"publishedAt":19},"01M7FEDNGFEMFWD1MSEDR5H7CY","bio-admet-prediction","Predicts ADMET properties using ADMETlab 3.0 (119 platform features, including 77 prediction models with modeled-endpoint uncertainty), ADMET-AI, DeepChem MolNet, and chemprop D-MPNN with explicit handling of OECD QSAR principles, applicability domain assessment, calibration, hERG\u002FCYP\u002FAMES endpoints, and PAINS \u002F Lipinski \u002F Ro5 \u002F Veber \u002F BBB druglikeness filters. Use when filtering compounds for drug-likeness, prioritizing leads by predicted safety, or building an in-house ADMET QSAR model.","https:\u002F\u002Fgithub.com\u002FCZX-LBJ\u002Fdeepseek-science\u002Fblob\u002Fmain\u002Fpackages\u002Fskill\u002Fskill-science\u002Fassets\u002Fsources\u002Fbioskills\u002Fchemoinformatics\u002Fadmet-prediction\u002FSKILL.md",[],[],{"skillNo":69,"name":70,"nameZh":6,"description":71,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":72,"categories":73,"keywords":74,"publishedAt":19},"01M7FED69GFB54FRF7AV5066DN","writing-systems-papers","Paragraph-level structural blueprint for 10-12 page systems papers targeting OSDI, SOSP, ASPLOS, NSDI, and EuroSys. Provides page allocation, paragraph templates, and writing patterns. Use when user says \"写系统论文\", \"systems paper structure\", \"OSDI paper\", \"SOSP paper\", or wants fine-grained structural guidance for a systems conference submission.","https:\u002F\u002Fgithub.com\u002FCZX-LBJ\u002Fdeepseek-science\u002Fblob\u002Fmain\u002Fpackages\u002Fskill\u002Fskill-science\u002Fassets\u002Fsources\u002Faris\u002Fskills\u002Fwriting-systems-papers\u002FSKILL.md",[],[],{"skillNo":76,"name":77,"nameZh":6,"description":78,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":79,"categories":80,"keywords":81,"publishedAt":19},"01M7FECVVDWZ0X0VEDTCE4B4DZ","wiki-enrich","Fill in the per-paper TODO sections of research-wiki\u002Fpapers\u002F\u003Cslug>.md pages that literature-ingest skills leave as bare scaffolds. Use when user says 'enrich wiki', 'fill paper TODOs', 'wiki body 補完', '把 paper 摘要寫進 wiki', 'research-wiki 自動填', or after a batch ingest that left papers\u002F as TODO scaffolds.","https:\u002F\u002Fgithub.com\u002FCZX-LBJ\u002Fdeepseek-science\u002Fblob\u002Fmain\u002Fpackages\u002Fskill\u002Fskill-science\u002Fassets\u002Fsources\u002Faris\u002Fskills\u002Fwiki-enrich\u002FSKILL.md",[],[],{"skillNo":83,"name":84,"nameZh":6,"description":85,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":86,"categories":87,"keywords":88,"publishedAt":19},"01M7FECD4BCJ2MTSBDDSE3ZAJ5","web-debug-search","Search GitHub, Stack Exchange, Chinese technical communities, official documentation, and general developer web sources for software errors, compatibility problems, API usage questions, and real-world workarounds. Use for debugging and discovery only; results are not paper-citation evidence.","https:\u002F\u002Fgithub.com\u002FCZX-LBJ\u002Fdeepseek-science\u002Fblob\u002Fmain\u002Fpackages\u002Fskill\u002Fskill-science\u002Fassets\u002Fsources\u002Faris\u002Fskills\u002Fweb-debug-search\u002FSKILL.md",[],[]]