Code Review Checklist

MIT License

Copyright (c) 2026 OpenAgentSkills Contributors

Permission is hereby granted, free of charge, to any person obtaining a copy
of this software and associated documentation files (the "Software"), to deal
in the Software without restriction, including without limitation the rights
to use, copy, modify, merge, publish, distribute, sublicense, and/or sell
copies of the Software, and to permit persons to whom the Software is
furnished to do so, subject to the following conditions:

The above copyright notice and this permission notice shall be included in all
copies or substantial portions of the Software.

THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM,
OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE
SOFTWARE.

Code Review Checklist

What this skill does

This skill directs the agent to perform a structured code review using a proven checklist. It catches bugs, security issues, performance problems, and readability concerns that are easy to miss in a quick scan. The output is a prioritized list of findings — not vague suggestions, but specific line-level feedback with clear explanations.

Use this before merging a PR, after a refactor, or any time you want a second opinion on code quality.

How to use

Claude Code / Cline

Copy this file to .agents/skills/code-review-checklist/SKILL.md in your project root.

Then ask the agent:

• *"Review this file using the Code Review Checklist skill."*
• *"Run the Code Review Checklist skill on my PR diff."*

Cursor

Add the contents of the "Prompt / Instructions" section below to your .cursorrules file, then open the file you want reviewed and ask Cursor to review it.

Codex

Paste the code and the instructions from the section below into the Codex chat.

The Prompt / Instructions for the Agent

When asked to review code, work through each section of this checklist in order. For every issue found, output:

• Severity: critical | major | minor | nit
• Location: file name and line number(s)
• Issue: one-sentence description
• Suggestion: what to do instead

---

1. Correctness

• [ ] Does the logic match the intended behavior (based on function name, comments, or context)?
• [ ] Are there off-by-one errors in loops or array accesses?
• [ ] Are all code paths handled, including edge cases (empty input, null, zero, negative numbers)?
• [ ] Are async operations properly awaited? Could there be race conditions?
• [ ] Are errors caught and handled appropriately, or silently swallowed?

2. Security

• [ ] Is any user input used in SQL queries, shell commands, or file paths without sanitization?
• [ ] Are secrets, API keys, or passwords hardcoded or logged?
• [ ] Are authentication and authorization checks in place before sensitive operations?
• [ ] Is output properly escaped before being rendered in HTML (XSS)?
• [ ] Are dependencies pinned to known-good versions?

3. Performance

• [ ] Are there any N+1 query patterns (a query inside a loop)?
• [ ] Is expensive work (network calls, heavy computation) done unnecessarily on every request?
• [ ] Are large data structures being copied when a reference would suffice?
• [ ] Are there obvious opportunities for caching or memoization?

4. Readability & Maintainability

• [ ] Are variable and function names descriptive and consistent with the surrounding codebase?
• [ ] Is any logic complex enough to warrant a comment but currently unexplained?
• [ ] Are there magic numbers or strings that should be named constants?
• [ ] Is the function doing more than one thing (violates single-responsibility)?
• [ ] Is there dead code, commented-out code, or TODO comments that should be resolved?

5. Tests

• [ ] Are the critical paths covered by tests?
• [ ] Do the tests actually assert meaningful behavior, or just that no error was thrown?
• [ ] Are edge cases (empty input, error states) tested?

---

After completing the checklist, output a summary with:

1. Total findings by severity

2. The top 3 issues to fix first (highest severity / most impactful)

3. An overall assessment: Approve | Approve with minor fixes | Needs changes

If no issues are found in a section, write "✓ No issues found" for that section rather than skipping it.

Example

Input to Agent:

> "Use the Code Review Checklist skill on this function:

>

> `typescript

> async function getUser(id: string) {

> const query = SELECT * FROM users WHERE id = '${id}';

> const result = await db.raw(query);

> return result.rows[0];

> }

> `"

Output from Agent:

Correctness

• Major | Line 3 | No null check on result.rows[0] — returns undefined if user not found, which may cause downstream errors. | Return null explicitly or throw a NotFoundError.

Security

• Critical | Line 2 | SQL injection vulnerability — id is interpolated directly into the query string. | Use a parameterized query: db.raw('SELECT * FROM users WHERE id = ?', [id]).

Performance

• Minor | Line 2 | SELECT * fetches all columns even if only a few are needed. | Specify the columns you actually use.

Summary

• 1 critical, 1 major, 1 minor
• Top issues: SQL injection (critical), missing null check (major), over-fetching columns (minor)
• Assessment: Needs changes

Notes

• If the diff is large, focus on files with the most business logic first.
• For nitpicks (formatting, naming), group them at the end rather than interspersing them with real issues.
• This skill does not run linters or static analysis — it is a reasoning-based review. For automated checks, run your existing lint/test pipeline first.