Skip to content
Get packs

Convex Code Reviewer

Reviews Convex backend code for auth, validator, performance and pattern problems before shipping.

Source Download skill
Repository
openclaw/clawhub
Stars
9,546
Licence
MIT
Files
1 file

What it does

When installed, the agent reviews code in a Convex project's convex/ folder using a fixed checklist. It checks security and authorization first (are users authenticated, is ownership verified, are scheduled jobs pointed at internal functions), then performance (indexed queries instead of full-table filters, no Date.now() in queries, bounded collect calls), then code quality (argument and return validators, no any types, awaited promises). Findings come back grouped as Critical, Important or Suggestion, each with a reason and a suggested fix.

Great for

  • Developers building an app on the Convex backend who want a second pair of eyes before deploying
  • Teams doing a security or authorization audit of existing Convex functions
  • Developers new to Convex who want to learn its common anti-patterns

Skip it if

  • Projects that do not use Convex, since the checklist is specific to Convex APIs and would produce irrelevant findings
  • Replacing a human security audit for production systems handling sensitive data, because it is a checklist-based review and can miss issues

Try asking

  • “Review the functions in convex/ before I ship this release”
  • “Audit my Convex mutations for missing auth checks and ownership verification”
  • “Check my Convex queries for performance problems like .filter() and unbounded .collect()”
  • “Look over my schema and validators and tell me what needs args or returns validators”
Step-by-step guide: how to use Convex Code ReviewerInstall, first run, examples, tips and FAQs

Install in 3 steps

  1. 1

    Download the skill

    The ZIP contains the 1 original files, the MIT licence and an attribution note.

    convex-reviewer.zip
  2. 2

    Add it to your agent

    Claude Code (all projects)

    mkdir -p ~/.claude/skills && unzip convex-reviewer.zip -d ~/.claude/skills/

    Claude Code / Agent SDK (this project only)

    mkdir -p .claude/skills && unzip convex-reviewer.zip -d .claude/skills/

    Claude apps (web/desktop): Settings → Capabilities → Skills → Upload skill, and choose the ZIP.

  3. 3

    Ask for the task

    No special command needed — the agent reads the skill's description and loads it when your request matches. Try one of the prompts above.

Get better results

  • Point the agent at a specific file or folder (for example convex/messages.ts) to get a focused review rather than a whole-project pass
  • Ask it to explain each finding in your own context, since the skill is designed to justify every suggested change
  • Treat its output as a first-pass checklist and still run your own tests and a human review for anything security-sensitive

What's inside

  • SKILL.md

Redistributed unmodified from openclaw/clawhub under the MIT licence. The assessment was drafted with AI from the skill's files on 2026-10-29. Skills can instruct your agent to run code — read SKILL.md and any scripts before installing.