Review Loop

MANDATORY FIRST ACTION

STOP. Before ANY other tool call, you MUST call TaskCreate.

If you're about to:

• Run setup.sh → STOP. TaskCreate first.
• Dispatch reviewer → STOP. TaskCreate first.
• Read any file → STOP. TaskCreate first.

Step 1: Create Tasks

First, check if code-simplifier:code-simplifier subagent is available in your Task tool's subagent_type list.

If code-simplifier IS available:

TaskCreate(subject: "Simplify code", description: "Run code-simplifier before review", activeForm: "Simplifying code")
TaskCreate(subject: "Iteration 1: Review", description: "Review and fix", activeForm: "Running iteration 1", metadata: {"iteration": 1, "of": 4})
TaskCreate(subject: "Iteration 2: Review", description: "Review and fix", activeForm: "Running iteration 2", metadata: {"iteration": 2, "of": 4})
TaskCreate(subject: "Iteration 3: Review", description: "Review and fix", activeForm: "Running iteration 3", metadata: {"iteration": 3, "of": 4})
TaskCreate(subject: "Iteration 4: Review", description: "Review and fix", activeForm: "Running iteration 4", metadata: {"iteration": 4, "of": 4})

Then set dependencies (ITER1 blocked by SIMPLIFY):

TaskUpdate(taskId: ITER1, addBlockedBy: [SIMPLIFY])
TaskUpdate(taskId: ITER2, addBlockedBy: [ITER1])
TaskUpdate(taskId: ITER3, addBlockedBy: [ITER2])
TaskUpdate(taskId: ITER4, addBlockedBy: [ITER3])
TaskUpdate(taskId: SIMPLIFY, status: "in_progress")

If code-simplifier is NOT available:

TaskCreate(subject: "Iteration 1: Review", description: "Review and fix", activeForm: "Running iteration 1", metadata: {"iteration": 1, "of": 4})
TaskCreate(subject: "Iteration 2: Review", description: "Review and fix", activeForm: "Running iteration 2", metadata: {"iteration": 2, "of": 4})
TaskCreate(subject: "Iteration 3: Review", description: "Review and fix", activeForm: "Running iteration 3", metadata: {"iteration": 3, "of": 4})
TaskCreate(subject: "Iteration 4: Review", description: "Review and fix", activeForm: "Running iteration 4", metadata: {"iteration": 4, "of": 4})

Then set dependencies and start:

TaskUpdate(taskId: ITER2, addBlockedBy: [ITER1])
TaskUpdate(taskId: ITER3, addBlockedBy: [ITER2])
TaskUpdate(taskId: ITER4, addBlockedBy: [ITER3])
TaskUpdate(taskId: ITER1, status: "in_progress")

Report to user: "code-simplifier plugin not installed. Recommended: claude plugin install code-simplifier@anthropic-official from https://github.com/anthropics/claude-plugins-official"

CHECKPOINT: Have you created tasks? If NO → do it now. If YES → continue.

Review Loop

You are an ORCHESTRATOR. You dispatch subagents. You do NOT touch code.

Process

digraph review_loop {
    rankdir=TB;
    "TaskCreate" [shape=box, style=bold];
    "Get REVIEW_DIR/TARGET_BRANCH" [shape=box];
    "code-simplifier available?" [shape=diamond];
    "Dispatch code-simplifier" [shape=box];
    "TaskList → find unblocked" [shape=box];
    "TaskGet → read iteration metadata" [shape=box];
    "Dispatch reviewer" [shape=box];
    "Read findings + triage" [shape=box];
    "Dispatch fix subagent per issue" [shape=box];
    "CHECKPOINT" [shape=doubleoctagon, style=bold, color=red];
    "TaskUpdate completed" [shape=box];
    "iteration < of?" [shape=diamond];
    "Commit" [shape=box];

    "TaskCreate" -> "Get REVIEW_DIR/TARGET_BRANCH";
    "Get REVIEW_DIR/TARGET_BRANCH" -> "code-simplifier available?";
    "code-simplifier available?" -> "Dispatch code-simplifier" [label="yes"];
    "code-simplifier available?" -> "TaskList → find unblocked" [label="no"];
    "Dispatch code-simplifier" -> "TaskList → find unblocked";
    "TaskList → find unblocked" -> "TaskGet → read iteration metadata";
    "TaskGet → read iteration metadata" -> "Dispatch reviewer";
    "Dispatch reviewer" -> "Read findings + triage";
    "Read findings + triage" -> "Dispatch fix subagent per issue";
    "Dispatch fix subagent per issue" -> "CHECKPOINT";
    "CHECKPOINT" -> "TaskUpdate completed";
    "TaskUpdate completed" -> "iteration < of?";
    "iteration < of?" -> "TaskList → find unblocked" [label="yes, continue"];
    "iteration < of?" -> "Commit" [label="no, done"];
}

Step 2: Get Config

If args provided, use them. Otherwise:

~/.claude/plugins/cache/onsails-cc/review-loop/*/skills/review-loop/scripts/setup.sh

Step 2.5: Code Simplification (if available)

If code-simplifier:code-simplifier subagent is available, dispatch it to simplify code changes between current branch and TARGET_BRANCH:

Task(subagent_type: "code-simplifier:code-simplifier",
     prompt: "Simplify code changes between HEAD and ${TARGET_BRANCH}. Focus only on files modified in this branch.")
TaskUpdate(taskId: SIMPLIFY, status: "completed")

Step 3: Each Iteration

1. TaskList → find first unblocked iteration task

2. TaskGet(taskId: CURRENT) → read metadata.iteration and metadata.of

3. TaskUpdate(taskId: CURRENT, status: "in_progress")

4. Dispatch reviewer:

   Task(subagent_type: "review-loop:local-reviewer",
        prompt: "OUTPUT: ${REVIEW_DIR}/iterN.md\nTARGET: ${TARGET_BRANCH}")
   

   Reviewer returns a summary. Display it to user.

5. Read findings and triage (YOU do this, not a subagent):

   Read(file_path: "${REVIEW_DIR}/iterN.md")
   

   Parse the ## Findings section. For each issue under Critical/Major:

   • Create a fix task: TaskCreate(subject: "Fix: [title]", description: "...", activeForm: "Fixing [title]")
   • Block next iteration: TaskUpdate(taskId: NEXT_ITER, addBlockedBy: [fix_task_ids])

   Display triage table:

   | # | Severity | File:Line | Issue | Action |
   |---|----------|-----------|-------|--------|
   | 1 | critical | foo.rs:42 | SQL injection | FIX |
   | 2 | major    | bar.rs:15 | Race condition | FIX |
   | 3 | minor    | baz.rs:99 | Unused import | SKIP |
   

   Skip rules: Skip minor/suggestion severity. Skip issues that are clearly false positives.

6. Dispatch ONE fix subagent per issue (YOU dispatch these directly):

   For each fix task, sequentially:

   TaskUpdate(taskId: FIX_TASK, status: "in_progress")
   Task(subagent_type: "general-purpose",
        description: "Fix: [issue title]",
        prompt: "Fix this specific issue:
   
   File: [file]:[line]
   Issue: [description]
   Severity: [severity]
   
   Make minimal changes to fix ONLY this issue.
   Do not fix other issues. Do not refactor surrounding code.
   Run relevant tests after fixing if test infrastructure exists.")
   TaskUpdate(taskId: FIX_TASK, status: "completed")
   

   Each Task call = ONE fix. Never batch multiple fixes into one prompt. Wait for each subagent to complete before dispatching next.

7. TaskUpdate(taskId: CURRENT, status: "completed")

CHECKPOINT (after all fixes dispatched)

STOP HERE. All fix subagents have completed. You MUST now check iteration progress:

Current iteration: ${metadata.iteration}
Total iterations: ${metadata.of}

Is ${metadata.iteration} < ${metadata.of}?
  YES → Go to Step 3.1 (TaskList → find next unblocked)
  NO  → Go to Step 4 (Completion)

DO NOT STOP after fixes complete. Fixing is a sub-step, not the end of the loop.

WHY 4+ iterations are mandatory:

• Reviewers find different issues on different passes
• Fixes may introduce new problems
• Context builds across iterations
• First pass often misses subtle issues

NEVER stop early because:

• "No issues found" → Reviewer may find different issues next pass
• "All false-positives" → Next iteration reviews with fresh perspective
• "Code looks clean" → Run all iterations anyway

Step 4: Completion

After 4+ iterations with no critical/major:

1. Mark final iteration task completed:

   TaskUpdate(taskId: ITER4, status: "completed")
   

2. Verify all tasks completed:

   TaskList  → All iteration tasks should show "completed"
   

3. Commit changes:

   git add -A && git commit -m "fix: address review issues (N iterations)"
   

Do NOT leave tasks in pending/in_progress state.

Rationalization Table

Red Flags - STOP IMMEDIATELY

If you catch yourself:

• Dispatching reviewer without tasks created → STOP
• Running setup.sh as first action → STOP
• Using Read/Edit/Grep on code → STOP
• Fixing issues directly (Edit/Write on code) → STOP
• Batching multiple fixes into one subagent → STOP
• Asking permission → STOP
• Skipping code-simplifier without checking availability → STOP
• Stopping before iteration 4 because "no issues" → STOP
• Skipping iterations because "all false-positives" → STOP
• Ending response after fixes complete → STOP (go to CHECKPOINT)
• Not checking metadata.iteration after fixes → STOP (read the task, check the count)

All mean: You violated the skill. Go back and follow it exactly.

Iron Rules

1. TaskCreate BEFORE anything else
2. Check code-simplifier availability, run if present
3. MINIMUM 4 review iterations
4. ONLY Task tool on code (dispatch subagents)
5. SEQUENTIAL iterations
6. ONE fix subagent per issue — never batch
7. YOU read findings and triage — no intermediary agent
8. Never ask permission