auto-review-loop

name	auto-review-loop
description	Autonomous multi-round research review loop. Repeatedly reviews via external reviewer script, 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.
argument-hint	["topic-or-scope"]
allowed-tools	Bash(*), Read, Grep, Glob, Write, Edit, Agent, Skill

Auto Review Loop: Autonomous Research Improvement

Autonomously iterate: review → implement fixes → re-review, until the external reviewer gives a positive assessment or MAX_ROUNDS is reached.

Context: $ARGUMENTS

Constants

• MAX_ROUNDS = 4
• POSITIVE_THRESHOLD: score >= 6/10, or verdict contains "accept", "sufficient", "ready for submission"
• REVIEW_DOC: AUTO_REVIEW.md in project root (cumulative log)
• REVIEWER_MODEL: the configured reviewer model, invoked via reviewer_client.py script
• 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.

💡 Override: /auto-review-loop "topic" — human checkpoint: true

State Persistence (Compact Recovery)

Long-running loops may hit the context window limit, triggering automatic compaction. To survive this, persist state to REVIEW_STATE.json after each round:

{
  "round": 2,
  "status": "in_progress",
  "last_score": 5.0,
  "last_verdict": "not ready",
  "pending_experiments": ["screen_name_1"],
  "timestamp": "2026-03-13T21:00:00"
}

Write this file at the end of every Phase E (after documenting the round). Overwrite each time — only the latest state matters.

On completion (positive assessment or max rounds), set "status": "completed" so future invocations don't accidentally resume a finished loop.

Workflow

Initialization

1. Check for REVIEW_STATE.json in project root:
   • If it does not exist: fresh start (normal case, identical to behavior before this feature existed)
   • If it exists AND status is "completed": fresh start (previous loop finished normally)
   • If it exists AND status is "in_progress" AND timestamp is older than 24 hours: fresh start (stale state from a killed/abandoned run — delete the file and start over)
   • If it exists AND status is "in_progress" AND timestamp is within 24 hours: resume
     • Read the state file to recover round, last_score, pending_experiments
     • Read AUTO_REVIEW.md to restore full context of prior rounds
     • If pending_experiments is non-empty, check if they have completed (e.g., check screen sessions)
     • Resume from the next round (round = saved round + 1)
     • Log: "Recovered from context compaction. Resuming at Round N."
2. Read project narrative documents, memory files, and any prior review documents
3. Read recent experiment results — explicitly check these files in order:
   • experiment_results.md (from experiment-bridge, contains structured results summary)
   • figures/experiment_data.json (consolidated raw experiment data)
   • RESULTS.md (from comp-code, if competition workflow)
   • Output directories: results/, outputs/, logs/
4. Identify current weaknesses and open TODOs from prior reviews
5. Initialize round counter = 1 (unless recovered from state file)
6. Create/update AUTO_REVIEW.md with header and timestamp

Loop (repeat up to MAX_ROUNDS)

Phase A: Review

Send comprehensive context to the external reviewer via reviewer_client.py.

If the reviewer script fails (API key not configured): perform the review yourself using your own critical analysis capabilities. Act as a senior ML reviewer (NeurIPS/ICML level) and score the work honestly. The loop can still function without external review, though cross-model review is preferred for objectivity.

When the reviewer script is available, use:

cat << 'REVIEW_EOF' > _review_prompt.txt
[Round N/MAX_ROUNDS of autonomous review loop]

[Full research context: claims, methods, results, known weaknesses]
[Changes since last round, if any]

Please act as a senior ML reviewer (NeurIPS/ICML level).

1. Score this work 1-10 for a top venue
2. List remaining critical weaknesses (ranked by severity)
3. For each weakness, specify the MINIMUM fix (experiment, analysis, or reframing)
4. State clearly: is this READY for submission? Yes/No/Almost

Be brutally honest. If the work is ready, say so clearly.
REVIEW_EOF
PYTHON=$(command -v python3 2>/dev/null || command -v python 2>/dev/null)
$PYTHON "$REVIEWER_SCRIPT" --prompt-file _review_prompt.txt --thread-file _reviewer_thread.json

If this is round 2+, use the same _reviewer_thread.json to maintain conversation context (对话历史通过 _reviewer_thread.json 自动保存).

Phase B: Parse Assessment

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.

Then extract structured fields:

• Score (numeric 1-10)
• Verdict ("ready" / "almost" / "not ready")
• Action items (ranked list of fixes)

STOP CONDITION: If score >= 6 AND verdict contains "ready" or "almost" → stop loop, document final state.

Human Checkpoint (if enabled)

Skip this step entirely if HUMAN_CHECKPOINT = false.

When HUMAN_CHECKPOINT = true, present the review results. In interactive mode, wait for user input. In non-interactive mode (claude -p), treat as "go" and proceed with all suggested fixes automatically. Log: "HUMAN_CHECKPOINT: non-interactive mode, auto-proceeding with all fixes."

📋 Round N/MAX_ROUNDS review complete.

Score: X/10 — [verdict]
Top weaknesses:
1. [weakness 1]
2. [weakness 2]
3. [weakness 3]

Suggested fixes:
1. [fix 1]
2. [fix 2]
3. [fix 3]

Options:
- Reply "go" or "continue" → implement all suggested fixes
- Reply with custom instructions → implement your modifications instead
- Reply "skip 2" → skip fix #2, implement the rest
- Reply "stop" → end the loop, document current state

Wait for the user's response. Parse their input:

• Approval ("go", "continue", "ok", "proceed"): proceed to Phase C with all suggested fixes
• Custom instructions (any other text): treat as additional/replacement guidance for Phase C. Merge with reviewer suggestions where appropriate
• Skip specific fixes ("skip 1,3"): remove those fixes from the action list
• Stop ("stop", "enough", "done"): terminate the loop, jump to Termination

Feishu Notification (if configured)

After parsing the score, check if ~/.claude/feishu.json exists and mode is not "off":

• Send a review_scored notification: "Round N: X/10 — [verdict]" with top 3 weaknesses
• If HUMAN_CHECKPOINT=true and verdict is "almost": send as checkpoint, wait for user reply on whether to continue or stop. In non-interactive mode (claude -p), auto-continue.
• If config absent or mode off: skip entirely (no-op)

Phase C: Implement Fixes (if not stopping)

For each action item (highest priority first):

1. Code changes: Write/modify experiment scripts, model code, analysis scripts
2. Run experiments: Deploy to GPU server via SSH + screen/tmux
3. Analysis: Run evaluation, collect results, update figures/tables
4. Documentation: Update project notes and review document

Prioritization rules:

• Skip fixes requiring excessive compute (flag for manual follow-up)
• Skip fixes requiring external data/models not available
• Prefer reframing/analysis over new experiments when both address the concern
• Always implement metric additions (cheap, high impact)

Phase D: Wait for Results

If experiments were launched:

• Monitor remote sessions for completion
• Collect results from output files and logs

Phase D2: Recompile + Quality Check

⛔ MANDATORY after every round of fixes. Do NOT skip this step — the user expects a compilable PDF after each round.

1. Pre-compile cleanup: Run compile_utils.sh to auto-fix common issues:

if [ -f "_utils/compile_utils.sh" ]; then
    bash _utils/compile_utils.sh paper/
elif [ -f "skills/shared-scripts/compile_utils.sh" ]; then
    bash skills/shared-scripts/compile_utils.sh paper/
fi

2. Compile (auto-detect engine):

cd paper/
# Detect engine
if grep -q 'ctex\|xelatex\|xeCJK\|fontspec' main.tex 2>/dev/null; then
    ENGINE=xelatex
else
    ENGINE=pdflatex
fi
$ENGINE -interaction=nonstopmode main.tex
bibtex main 2>/dev/null
$ENGINE -interaction=nonstopmode main.tex
$ENGINE -interaction=nonstopmode main.tex
cd ..

3. Post-compile check: Run compile_check.sh and writing_check.sh:

bash _utils/compile_check.sh paper/ 2>/dev/null || bash skills/shared-scripts/compile_check.sh paper/ 2>/dev/null
bash _utils/writing_check.sh paper/ 2>/dev/null || bash skills/shared-scripts/writing_check.sh paper/ 2>/dev/null

4. Verify PDF: Confirm paper/main.pdf exists and is non-trivial (>100KB):

if [ -f paper/main.pdf ] && [ $(wc -c < paper/main.pdf) -gt 100000 ]; then
    echo "✅ PDF compiled successfully ($(wc -c < paper/main.pdf) bytes)"
    cp paper/main.pdf "paper/main_round${ROUND}.pdf"
else
    echo "⛔ PDF compilation failed — fix errors before proceeding"
fi

If compilation fails, fix the LaTeX errors (check paper/main.log) and recompile before moving to Phase E. Do NOT proceed with a broken PDF.

Phase E: Document Round

Append to AUTO_REVIEW.md:

## Round N (timestamp)

### Assessment (Summary)
- Score: X/10
- Verdict: [ready/almost/not ready]
- Key criticisms: [bullet list]

### Reviewer Raw Response

<details>
<summary>Click to expand full reviewer response</summary>

[Paste the COMPLETE raw response from the external reviewer here — verbatim, unedited.
This is the authoritative record. Do NOT truncate or paraphrase.]

</details>

### Actions Taken
- [what was implemented/changed]

### Results
- [experiment outcomes, if any]

### Status
- [continuing to round N+1 / stopping]

Write REVIEW_STATE.json with current round, score, verdict, and any pending experiments.

Increment round counter → back to Phase A.

Termination

When loop ends (positive assessment or max rounds):

1. Update REVIEW_STATE.json with "status": "completed"
2. Write final summary to AUTO_REVIEW.md
3. Final compile + quality gate:

   # Pre-compile cleanup
   bash _utils/compile_utils.sh paper/ 2>/dev/null || bash skills/shared-scripts/compile_utils.sh paper/ 2>/dev/null
   # Compile
   cd paper/
   ENGINE=pdflatex
   grep -q 'ctex\|xelatex\|xeCJK\|fontspec' main.tex 2>/dev/null && ENGINE=xelatex
   $ENGINE -interaction=nonstopmode main.tex
   bibtex main 2>/dev/null
   $ENGINE -interaction=nonstopmode main.tex
   $ENGINE -interaction=nonstopmode main.tex
   cd ..
   # Post-compile check
   bash _utils/compile_check.sh paper/ 2>/dev/null || bash skills/shared-scripts/compile_check.sh paper/ 2>/dev/null
   bash _utils/writing_check.sh paper/ 2>/dev/null || bash skills/shared-scripts/writing_check.sh paper/ 2>/dev/null
   # Verify
   [ -f paper/main.pdf ] && echo "✅ Final PDF: $(wc -c < paper/main.pdf) bytes" || echo "⛔ Final PDF missing"
   

   ⛔ If final PDF is missing or <100KB, fix compilation errors before finishing.
4. Write NARRATIVE_REPORT.md in the project root — a comprehensive research narrative document containing:
   • Problem description and motivation
   • Methodology overview
   • Experimental setup and results (with key metrics from collected data)
   • Claims-evidence mapping (which experiments support which claims)
   • Known limitations and remaining weaknesses
   • This document serves as the primary input for /paper-plan and /paper-write in Workflow 3.
5. Update project notes with conclusions
6. Write method/pipeline description to AUTO_REVIEW.md under a ## Method Description section — a concise 1-2 paragraph description of the final method, its architecture, and data flow. This serves as input for /paper-illustration in Workflow 3 (so it can generate architecture diagrams automatically).

Key Rules

• Large file handling: For long output files (reports, reviews, etc.), ALWAYS use Bash heredoc (cat << 'EOF' > file and cat << 'EOF' >> file) to write in chunks instead of the Write tool. The Write tool may silently fail on large content due to output token limits, causing empty parameters. Do NOT use the Write tool for files longer than ~150 lines.

• ALWAYS use the same _reviewer_thread.json across rounds to maintain conversation context

• 对话历史通过 _reviewer_thread.json 自动保存

• Be honest — include negative results and failed experiments

• Do NOT hide weaknesses to game a positive score

• Implement fixes BEFORE re-reviewing (don't just promise to fix)

• If an experiment takes > 30 minutes, launch it and continue with other fixes while waiting

• Document EVERYTHING — the review log should be self-contained

• Update project notes after each round, not just at the end

Prompt Template for Round 2+

cat << 'REVIEW_EOF' > _review_prompt.txt
[Round N update]

Since your last review, we have:
1. [Action 1]: [result]
2. [Action 2]: [result]
3. [Action 3]: [result]

Updated results table:
[paste metrics]

Please re-score and re-assess. Are the remaining concerns addressed?
Same format: Score, Verdict, Remaining Weaknesses, Minimum Fixes.
REVIEW_EOF
PYTHON=$(command -v python3 2>/dev/null || command -v python 2>/dev/null)
$PYTHON "$REVIEWER_SCRIPT" --prompt-file _review_prompt.txt --thread-file _reviewer_thread.json