Claude Code Code Review Prompt Template
You get a prioritized list of real problems in the code, each tied to a line and paired with a concrete fix, without a pile of style nitpicks. This version is set up for Claude Code: name the files, commands and tests involved. Claude Code can read files and run commands in your project, asking permission as it goes. Below you'll find the prompt, how to run it in Claude Code, a worked example and answers to common questions.
Review the code below like a senior engineer on my team would. ## Context Language and framework: [e.g. TypeScript, Node 20, Express] What this code does: [one sentence] What I want you to focus on: [e.g. correctness, readability, edge cases] ## Code [paste the code or diff] ## Instructions - Look for bugs first, then edge cases, then readability. - Point to the specific line or function for every issue. - For each issue, show the corrected code, not just a description. - Skip pure formatting comments unless they hide a real problem. - If something looks wrong but might be intentional, say so and ask. ## Output format A list grouped by severity: - Critical: will break or corrupt data - Should fix: likely bugs, missing edge cases, confusing logic - Optional: small readability improvements End with a one-line overall verdict: ready to merge, merge after fixes, or needs rework. Context - Files involved: [list the files, or reference them in the chat] - How I build and test: [e.g. npm test]
How to use this prompt
- 1
Open Claude Code
Run claude in your project directory and paste the prompt.
- 2
Give it context
Name the files, commands and tests involved. Claude Code can read files and run commands in your project, asking permission as it goes.
- 3
Review and iterate
Approve or decline each edit and command, then reply in the same session to adjust.
- 4
Save what you repeat
Put project conventions (how to run tests, style rules, folders to avoid) in a CLAUDE.md file so every session starts with them.
Working in Claude Code
- •Claude Code works in your terminal, so tell it the exact commands for building and testing.
- •Ask it to explain its plan before large edits; you can stop it if the plan is wrong.
- •Start a fresh session for an unrelated task so earlier context doesn't steer the new one.
Worked example
What you give it
Language: JavaScript, Node 20
What it does: returns the names for a list of user IDs.
Focus: correctness.
async function getUserNames(ids) {
const names = [];
ids.forEach(async (id) => {
const user = await db.users.find(id);
names.push(user.name);
});
return names;
}What a good answer looks like (shortened)
Critical
- forEach does not wait for async callbacks, so the function returns an empty array before any lookup finishes. Use Promise.all:
async function getUserNames(ids) {
const users = await Promise.all(ids.map((id) => db.users.find(id)));
return users.map((user) => user.name);
}
Should fix
- If an ID does not exist, db.users.find likely returns null and user.name throws. Decide whether to skip missing users (users.filter(Boolean)) or raise a clear error.
Optional
- Promise.all fires every query at once. For very large ID lists, a single query by ID list would be cheaper.
Verdict: needs rework.Tips for better results
- •Use the focus line in Context to steer the review. "Correctness and edge cases" gets very different feedback than "readability for new hires."
- •Paste a diff rather than a whole file when you are reviewing a pull request, and say so. The AI will then comment on the change instead of old code.
- •Start a fresh session for an unrelated task so earlier context doesn't steer the new one.
- •Avoid this: Asking for a review of an entire repository at once, which produces shallow, generic comments.
Get the free prompt pack
Our best coding prompts in one Markdown file, delivered by email. The templates here stay free either way — the pack just saves you the copying.
Variations
Pull request review
Replace the Code section with: ## Diff [paste the git diff] and add: Only comment on lines added or changed in this diff.
Team standards
Add this section: ## Our conventions: [paste your style rules or lint config] and flag only violations of these.
Quick pass
Change Output format to: The three most important issues only, one line each, with the fix.
Save it as a Retext snippet
Retext is a free text expander for Mac. Save this prompt once, then type ;codereview in any app (ChatGPT, Claude, your editor, a browser tab) and the full prompt is pasted in.
- Click Add to Retext. The prompt installs as a snippet with the trigger ;codereview.
- Change the trigger in Retext if you prefer a different one.
- Type the trigger wherever you want the prompt, then fill in the brackets.
Frequently asked questions
Can AI replace human code review?
It is good at catching mechanical mistakes like missed awaits, null checks and off-by-one errors. It does not know your product decisions or team history, so a person should still approve the change.
Where should I keep instructions I repeat every session?
In CLAUDE.md at the project root. Keep the prompt on this page for the specific task.
Why does the AI flag things that are intentional?
It only sees the code you paste. Add a line in Context explaining unusual choices, or keep the instruction asking it to flag possible intent instead of calling it a bug.
How do I reuse this prompt without copying it every time?
Click Add to Retext to save it as a snippet. Then type ;codereview in any Mac app, including your AI tool, and the full prompt is pasted in.
Related Templates
Get the free prompt pack
Our best coding prompts in one Markdown file, delivered by email. The templates here stay free either way — the pack just saves you the copying.