abacusai/abacusai-bot/code-review
Claude Code and agent skill by abacusai/abacusai-bot
- community
Checking that the SKILL.md file is still in the repository.
Review a diff or pull request for correctness, security, and clarity, and report findings that are worth someone's time. Use when asked to review code, check a change before merging, or look over a PR.
Install
npx vskill@latest install abacusai/abacusai-bot/code-reviewSKILL.md
Code review
A review is only useful if its findings are real. Ten nitpicks bury the one bug, and a confident false positive costs more time than saying nothing.
Read in this order
- The description. What is this supposed to do? A change that works but
solves a different problem is still wrong.
- The tests. They tell you what the author believes the contract is.
- The diff, in dependency order: data model, then logic, then callers.
- What is missing. The absent null check, the unhandled error path, the
caller that was not updated, the test that was not written.
What actually matters, in order
Correctness. Does it do what it claims? Off-by-one, wrong operator, inverted condition, missing await, mutation of a shared object, error swallowed by a bare except.
Edge cases. Empty, one, many. Null and undefined. Zero and negative. Unicode. Concurrent callers. What happens on the second call?
Security. Untrusted input reaching a query, a shell, a path, or a template. Secrets in code or logs. Authorisation checked at the right layer. Anything that lets a user address someone else's data.
Resource handling. Files and sockets closed on the error path too. Unbounded growth. A query inside a loop.
Clarity, but only where it will genuinely confuse the next reader. Naming that misleads is worth raising; naming you would have chosen differently is not.
Before you report a finding
Ask: what input makes this fail, and what happens? If you cannot answer concretely, you have a hunch, not a finding. Either verify it or drop it.
Trace the actual code path rather than pattern-matching. "This looks like a SQL injection" is worth ten seconds of checking whether the value is parameterised before you say it.
How to say it
- Lead with the consequence: "Returns 500 for an empty cart" beats *"consider
guarding this"*.
- Separate blocking from optional, explicitly.
- Praise is not padding: if something is done well, saying so tells the author
which instincts to keep.
- If the whole approach is wrong, say that first. Line comments on code that
should not exist waste everyone's time.
Security
- Security scanPassed, 90 of 100Pattern checks on the skill's files.
- Model reviewPassed
Quality evaluation
- Task evaluationNot runNo evaluation yet. Run one in Skill Studio.
Source
- SKILL.md in the repositoryCheckingapps/desktop/resources/skills/code-review/SKILL.md
Popularity
- 7-day trend score
- 0
Not refreshed yet.
Works with
Any agent vskill installs to: 56 targets in all. These 39 have a page on this site.
- Claude Code
- Cursor
- GitHub Copilot
- Windsurf
- Codex
- Gemini CLI
- Cline
- Amp
- Roo Code
- Goose
- Continue
- Kilo Code
- Junie
- Trae
- OpenHands
- Qwen Code
- AI Agent
- Tabnine Chat
- Sourcegraph Cody
- Amazon Q
- Firebase Genkit
- Sweep
- Devin
- Phind
- Mentat
- Claude Desktop
- Aider
- GPT Pilot
- Plandex
- bolt.new
- v0
- Replit Agent
- Anse
- TypingMind
- BetterChatGPT
- LibreChat
- Warp AI
- Pieces
- Fig AI
Embed badge
[](https://verified-skill.com/skills/abacusai/abacusai-bot/code-review)