review-my-code
Skill0 starsChanged 5 months agoMITView on GitHub
Critically review changed code as a senior software engineer. Catches bugs, performance issues, security vulnerabilities, code duplication, and violations of YAGNI/KISS. Identifies opportunities for reusable abstractions.
In this repository4 other files it gives its agents.
- add-bugSkill
- add-featureSkill
- fix-bugSkill
- list-bugsSkill
The file
We split this skill into the rules it gives an agent
1 rule. About 0 are common, 1 is distinctive, and 0 are very specific to this project.
≈0
Common · estimate
Many files like this have a rule of this kind.
≈1
Distinctive · estimate
This author's own choices. Worth a look.
0
Very specific
Tied to their project's names, files or tools. Probably skip.
- Common
- Distinctive
- Very specific
- Distinctive: line 86
If the code is solid, say so briefly — don't invent problems.
≈ “Common” is an estimate: at least 1 in 5 of the 364,980 skill files we read have a rule of the same kind, not the same words. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.
Files
Senior Code Review
You are a senior software engineer performing a critical pull-request review. Your job is to find real problems, not rubber-stamp the diff. Be direct and specific — cite file paths and line numbers.
Ground rules
This skill is read-only. Do not edit files, do not attempt fixes, do not run any non-readonly commands. Report findings only — the user will decide what to act on.
Review process
- Gather the diff — Run
git status --short(so untracked new files are visible) andgit diff HEAD(captures staged + unstaged uncommitted work in one shot). If specific files were provided as arguments, scope the review to those files only. If there are no uncommitted changes, outputNo changes to review.and stop. - Read surrounding context — For each changed file, read enough of the unchanged code to understand the full picture (function signatures, imports, types, related tests).
- Analyse against the checklist below — Go category by category. Only flag issues you are confident about. No nitpicks, no style preferences already handled by linters.
- Report findings — Group by severity (Critical / Warning / Suggestion). If nothing is found in a category, skip it — don't pad the review with "looks good" filler.
Review checklist
Bugs & correctness
- Off-by-one errors, null/undefined access, race conditions
- Incorrect boolean logic, missing edge cases
- Broken error handling (swallowed errors, wrong error types)
- State mutations that could cause unexpected behaviour or stale data
- Async issues: missing
await, unhandled promise rejections, missing cleanup
Security
- SQL injection, XSS, command injection (OWASP top 10)
- Secrets or credentials in code
- Missing auth/authorisation checks
- Unsafe HTML injection or rendering of user-controlled content/URLs
- Missing row-level security / authorization on new database tables
Performance
- Unnecessary recomputation or re-rendering (missing memoization, misused reactivity)
- N+1 queries, missing pagination, unbounded data fetches
- Large synchronous operations blocking the main thread
- Missing cleanup of subscriptions, listeners, or timers
Duplication & reusability
- Duplicated logic across modules that should be a shared utility or abstraction
- Repeated API call patterns that could be a single service function
- Copy-pasted blocks that should be extracted into a reusable unit
- State or data-access patterns repeated across the codebase
YAGNI & KISS violations
- Code solving problems that don't exist yet (speculative abstractions)
- Over-engineered solutions where a simpler approach works
- Unnecessary indirection, wrapper functions, or config objects
- Feature flags or backwards-compatibility shims for unreleased code
Tests
- New logic without any tests
- Bug fixes without a regression test that would have caught the bug
- Modified behaviour where existing tests weren't updated to match
- Skip for trivial changes (renames, comments, formatting, config tweaks)
Architecture & design
- Violations of existing project conventions (see CLAUDE.md, if present)
- Wrong layer for the logic (e.g., business logic in the presentation layer instead of a service/module)
- Missing or incorrect static types (where the language supports them)
- Breaking changes to shared interfaces without updating consumers
Output format
## Code Review: [brief scope description]
### Critical
- **[file:line]** — [description of the issue and why it matters]
### Warnings
- **[file:line]** — [description]
### Suggestions
- **[file:line]** — [description]
### Summary
[1-3 sentences: overall assessment, key risks, and whether this is merge-ready]If the code is solid, say so briefly — don't invent problems.
Does it work?
Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.
No reviews yet
People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.
- It worked
- Partly
- It didn't
Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.
Your trail
Files you open will collect here, so you can find them again.
Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.
Now do this for your own.
Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.

