@hustle-together/api-dev-tools 3.12.16 → 4.5.3

package/.claude/hooks/run-code-review.py

@@ -1,393 +0,0 @@
-#!/usr/bin/env python3
-"""
-Phase 11: AI Code Review Hook (Ralph Wiggum Loop Pattern)
-
-Triggers Greptile AI code review and LOOPS until all issues are fixed.
-This ensures code quality before proceeding to refactor phase.
-
-Hook Type: PostToolUse (triggers after tests pass)
-
-Ralph Wiggum Pattern:
-    1. Run Greptile review
-    2. If issues found → inject context for agent to fix
-    3. Agent fixes issues
-    4. Tests re-run → hook triggers again
-    5. Re-review with Greptile
-    6. Loop until clean OR max iterations
-    7. Emit <promise>REVIEW_CLEAN</promise>
-
-Environment Variables:
-    GREPTILE_API_KEY: Your Greptile API key (get from https://app.greptile.com)
-    GITHUB_TOKEN: GitHub Personal Access Token with repo access
-    CODE_REVIEW_ENABLED: Set to 'true' to enable (default: true)
-    CODE_REVIEW_MAX_ITERATIONS: Max review cycles (default: 5)
-
-Version: 2.0.0
-"""
-import os
-import sys
-import json
-import subprocess
-from pathlib import Path
-from datetime import datetime
-
-# Add lib directory to path for imports
-HOOK_DIR = Path(__file__).parent
-LIB_DIR = HOOK_DIR / "lib"
-sys.path.insert(0, str(LIB_DIR))
-
-try:
-    from greptile import (
-        is_configured,
-        review_changes,
-        get_review_summary,
-        format_review_for_display,
-        get_status
-    )
-    GREPTILE_AVAILABLE = True
-except ImportError:
-    GREPTILE_AVAILABLE = False
-
-# State file for tracking review loops
-REVIEW_STATE_FILE = ".claude/code-review-state.json"
-MAX_ITERATIONS = int(os.environ.get("CODE_REVIEW_MAX_ITERATIONS", "5"))
-
-
-def get_git_diff() -> tuple:
-    """Get the current git diff and changed files."""
-    try:
-        # Get list of changed files
-        files_result = subprocess.run(
-            ["git", "diff", "--name-only", "HEAD~1"],
-            capture_output=True,
-            text=True,
-            timeout=30
-        )
-        files_changed = files_result.stdout.strip().split("\n") if files_result.stdout else []
-
-        # Get full diff
-        diff_result = subprocess.run(
-            ["git", "diff", "HEAD~1"],
-            capture_output=True,
-            text=True,
-            timeout=30
-        )
-        diff_content = diff_result.stdout
-
-        return files_changed, diff_content
-    except (subprocess.TimeoutExpired, FileNotFoundError):
-        return [], ""
-
-
-def get_repo_info() -> tuple:
-    """Get repository owner and name from git remote."""
-    try:
-        result = subprocess.run(
-            ["git", "remote", "get-url", "origin"],
-            capture_output=True,
-            text=True,
-            timeout=10
-        )
-        if result.returncode == 0:
-            url = result.stdout.strip()
-            # Parse GitHub URL (handles both HTTPS and SSH)
-            if "github.com" in url:
-                if url.startswith("git@"):
-                    # SSH format: git@github.com:owner/repo.git
-                    parts = url.split(":")[-1].replace(".git", "").split("/")
-                else:
-                    # HTTPS format: https://github.com/owner/repo.git
-                    parts = url.replace(".git", "").split("/")[-2:]
-
-                if len(parts) >= 2:
-                    return parts[-2], parts[-1]
-    except (subprocess.TimeoutExpired, FileNotFoundError):
-        pass
-    return None, None
-
-
-def load_review_state() -> dict:
-    """Load code review loop state."""
-    state_file = Path.cwd() / REVIEW_STATE_FILE
-    if state_file.exists():
-        try:
-            return json.loads(state_file.read_text())
-        except (json.JSONDecodeError, IOError):
-            pass
-    return {
-        "iteration": 0,
-        "issues_found": [],
-        "status": "pending",
-        "started_at": None,
-        "last_review_at": None
-    }
-
-
-def save_review_state(state: dict):
-    """Save code review loop state."""
-    state_file = Path.cwd() / REVIEW_STATE_FILE
-    state_file.parent.mkdir(parents=True, exist_ok=True)
-    try:
-        state_file.write_text(json.dumps(state, indent=2))
-    except IOError:
-        pass
-
-
-def clear_review_state():
-    """Clear review state after successful completion."""
-    state_file = Path.cwd() / REVIEW_STATE_FILE
-    if state_file.exists():
-        try:
-            state_file.unlink()
-        except IOError:
-            pass
-
-
-def load_workflow_state() -> dict:
-    """Load current workflow state."""
-    state_file = Path.cwd() / ".claude" / "api-dev-state.json"
-    if state_file.exists():
-        try:
-            return json.loads(state_file.read_text())
-        except (json.JSONDecodeError, IOError):
-            pass
-    return {}
-
-
-def update_workflow_state_with_review(review_summary: dict, iteration: int):
-    """Update workflow state file with code review results."""
-    state_file = Path.cwd() / ".claude" / "api-dev-state.json"
-    state = load_workflow_state()
-
-    # Add or update code_review phase
-    if "phases" not in state:
-        state["phases"] = {}
-
-    state["phases"]["code_review"] = {
-        "status": "in_progress" if review_summary.get("issue_count", 0) > 0 else "complete",
-        "iteration": iteration,
-        "score": review_summary.get("score", 0),
-        "issues_found": review_summary.get("issue_count", 0),
-        "suggestions": review_summary.get("suggestion_count", 0),
-        "reviewed_at": datetime.now().isoformat()
-    }
-
-    try:
-        state_file.write_text(json.dumps(state, indent=2))
-    except IOError:
-        pass
-
-
-def should_run_review(hook_input: dict) -> bool:
-    """Determine if code review should run based on hook context."""
-    # Check if code review is enabled
-    if os.environ.get("CODE_REVIEW_ENABLED", "true").lower() == "false":
-        return False
-
-    tool_name = hook_input.get("tool_name", "")
-
-    # Run after tests pass (Phase 9/10)
-    if tool_name == "Bash":
-        tool_input = hook_input.get("tool_input", {})
-        command = tool_input.get("command", "")
-        tool_result = hook_input.get("tool_result", {})
-        stdout = tool_result.get("stdout", "")
-
-        # Check if tests just passed
-        if ("pnpm test" in command or "npm test" in command or "vitest" in command):
-            # Only run if tests passed (look for success indicators)
-            if "pass" in stdout.lower() or "✓" in stdout or "PASS" in stdout:
-                return True
-
-    # Also run if verify-after-green hook triggered
-    if "verify" in tool_name.lower():
-        return True
-
-    return False
-
-
-def format_issues_for_context(issues: list) -> str:
-    """Format issues as context for the agent to fix."""
-    if not issues:
-        return ""
-
-    lines = [
-        "",
-        "=" * 60,
-        "CODE REVIEW ISSUES TO FIX (Ralph Wiggum Loop)",
-        "=" * 60,
-        "",
-        "The following issues were found by Greptile code review.",
-        "Please fix ALL issues, then run tests again.",
-        "The review will re-run automatically after tests pass.",
-        "",
-        "ISSUES:",
-    ]
-
-    for i, issue in enumerate(issues, 1):
-        lines.append(f"  {i}. {issue}")
-
-    lines.extend([
-        "",
-        "After fixing all issues, run: pnpm test",
-        "Review will loop until all issues are resolved.",
-        "=" * 60,
-        ""
-    ])
-
-    return "\n".join(lines)
-
-
-def main():
-    """Main hook entry point with Ralph Wiggum loop pattern."""
-    # Read hook input
-    try:
-        hook_input = json.loads(sys.stdin.read())
-    except json.JSONDecodeError:
-        hook_input = {}
-
-    # Check if we should run
-    if not should_run_review(hook_input):
-        # Pass through - no review needed
-        print(json.dumps({"continue": True}))
-        return
-
-    # Check if Greptile is available and configured
-    if not GREPTILE_AVAILABLE:
-        print(json.dumps({
-            "continue": True,
-            "message": "Greptile library not found - skipping code review"
-        }))
-        return
-
-    if not is_configured():
-        status = get_status()
-        print(json.dumps({
-            "continue": True,
-            "message": f"Phase 11 Code Review skipped: {status['message']}"
-        }))
-        return
-
-    # Load current review loop state
-    review_state = load_review_state()
-
-    # Increment iteration
-    review_state["iteration"] += 1
-    iteration = review_state["iteration"]
-
-    # Check max iterations
-    if iteration > MAX_ITERATIONS:
-        print(json.dumps({
-            "continue": True,
-            "message": f"Code review max iterations ({MAX_ITERATIONS}) reached. Proceeding with warnings.\n\n<promise>REVIEW_CLEAN</promise>"
-        }))
-        clear_review_state()
-        return
-
-    # Track start time on first iteration
-    if iteration == 1:
-        review_state["started_at"] = datetime.now().isoformat()
-
-    review_state["last_review_at"] = datetime.now().isoformat()
-
-    # Get repository info
-    repo_owner, repo_name = get_repo_info()
-    if not repo_owner or not repo_name:
-        print(json.dumps({
-            "continue": True,
-            "message": "Could not determine repository info - skipping code review"
-        }))
-        return
-
-    # Get changes to review
-    files_changed, diff_content = get_git_diff()
-    if not diff_content:
-        print(json.dumps({
-            "continue": True,
-            "message": "No changes detected - code review complete.\n\n<promise>REVIEW_CLEAN</promise>"
-        }))
-        clear_review_state()
-        return
-
-    # Run the Greptile review
-    result = review_changes(
-        repo_owner=repo_owner,
-        repo_name=repo_name,
-        files_changed=files_changed,
-        diff_content=diff_content
-    )
-
-    if not result:
-        print(json.dumps({
-            "continue": True,
-            "message": "Greptile review failed - check API credentials"
-        }))
-        return
-
-    # Parse results
-    summary = get_review_summary(result)
-    update_workflow_state_with_review(summary, iteration)
-
-    # Format for display
-    display_output = format_review_for_display(summary)
-
-    # Check if issues found
-    issue_count = summary.get("issue_count", 0)
-    issues = summary.get("issues", [])
-
-    if issue_count == 0:
-        # All clean! Emit promise and proceed
-        review_state["status"] = "complete"
-        save_review_state(review_state)
-
-        output = f"""
-{display_output}
-
-================================================================================
-REVIEW LOOP COMPLETE (Iteration {iteration}/{MAX_ITERATIONS})
-================================================================================
-All code review checks passed!
-Proceeding to next phase.
-
-<promise>REVIEW_CLEAN</promise>
-"""
-        print(json.dumps({
-            "continue": True,
-            "message": output
-        }))
-        clear_review_state()
-        return
-
-    # Issues found - save state and block for fixes
-    review_state["status"] = "needs_fixing"
-    review_state["issues_found"] = issues
-    save_review_state(review_state)
-
-    # Format issues as context for agent
-    issues_context = format_issues_for_context(issues)
-
-    output = f"""
-{display_output}
-
-================================================================================
-REVIEW LOOP - ITERATION {iteration}/{MAX_ITERATIONS}
-================================================================================
-{issue_count} issue(s) found. Fix them and run tests again.
-Review will re-run automatically after tests pass.
-{issues_context}
-"""
-
-    # Block workflow - agent needs to fix issues
-    print(json.dumps({
-        "continue": False,  # Block until fixed
-        "message": output,
-        "review_score": summary.get("score", 0),
-        "issues_count": issue_count,
-        "iteration": iteration,
-        "action_required": True,
-        "next_action": "Fix the issues above, then run tests again"
-    }))
-
-
-if __name__ == "__main__":
-    main()