Prompt file imported from idealase/geeraff (.github/prompts/code-review.prompt.md). Copyright stays with the author.

Code Review

Dispatch this prompt to a review agent evaluating a PR or agent output.

Pull Request

{pr_url}

Issue

{issue_url}

Repository

{repo_name} — clone at ~/Desktop/Repos/{repo_name}

Summary of Changes

{change_summary}

Review Checklist

Correctness

• Solves the stated problem in the linked issue
• All acceptance criteria from the issue are met
• Edge cases handled (nulls, empty inputs, boundary values)
• Error handling is explicit, not swallowed
• No off-by-one errors in loops or slices

Security

• No secrets, tokens, or credentials in code or comments
• No eval(), dangerouslySetInnerHTML, or equivalent
• User input is validated and sanitized
• No new attack surface (XSS, injection, path traversal)
• Dependencies are from trusted sources

Architecture

• Follows existing patterns in the codebase
• No unnecessary abstractions or premature generalization
• Changes are in the correct layer (data / logic / presentation)
• No tight coupling introduced between unrelated modules
• File organization matches repo conventions

Tests

• Tests are meaningful — assert behavior, not implementation
• Tests would fail if the feature were removed
• Edge cases are tested, not just the happy path
• No flaky patterns (timers, network calls, random values)
• Test names describe the scenario, not the method

Performance

• No N+1 queries or unbounded loops
• No blocking calls on the main thread / event loop
• Large lists are paginated or virtualized
• No unnecessary re-renders (React) or recomputation

Common Agent Failure Modes

Watch specifically for these patterns in agent-generated code:

• Scope creep: Files modified that weren't in the issue's "Files to Modify"
• Phantom dependencies: npm install or pip install for packages not justified by the issue
• Test theater: Tests that always pass regardless of implementation
• Hallucinated APIs: Calling methods or endpoints that don't exist in the codebase
• Copy-paste drift: Code duplicated instead of extracting shared logic
• Incomplete error paths: Happy path works but errors crash or hang
• Stale imports: Importing from moved/renamed modules

Verdict Format

Respond with one of:

✅ PASS

All checklist items satisfied. Note any minor suggestions (non-blocking).

🔄 NEEDS CHANGES

For each required change, provide:

File: {path}
Line: {line_number}
Issue: {description of the problem}
Fix: {what should change}
Severity: blocking | important | minor

❌ REJECT

Fundamental problems requiring a rework. Explain why and reference the specific acceptance criteria that are unmet.

Context Files

{context_file_list}