qa-reviewer · first-fluke — Claude Code subagent: security
qa-reviewer — a code review agent that scans for OWASP vulnerabilities, performance bottlenecks, accessibility gaps and code quality issues. Runs before you merge and flags each finding with severity and remediation.
MIT ★ 1,472 updated 2026-10-29
/plugin marketplace add first-fluke/oh-my-agent
/plugin install oma@oh-my-agentWhat qa-reviewer does
qa-reviewer evaluates every change across four domains in priority order: security (OWASP Top 10), performance (N+1 queries, unnecessary re-renders, bundle bloat), accessibility (WCAG 2.2 AA), and code quality (naming, error handling, test coverage). It runs automated tooling — linters, type-checking, and dependency audits (npm audit, bandit, lighthouse where applicable) — then verifies each result by hand to rule out false positives.
Findings are grouped by severity: CRITICAL (block the merge), HIGH (fix before merging), MEDIUM (backlog it), LOW (nice-to-have). Every issue includes the file, line number, a clear description, and code to fix it. The agent reviews only; it does not touch your source.
Works in Claude Code and Codex. Needs access to your project dependencies to run audits.
What it helps with
- uncover OWASP security flaws before merging
- catch performance regressions: N+1 queries, redundant renders, bundle size
- validate code against WCAG 2.2 AA accessibility standards
- audit code quality: naming, exception handling, test gaps
Example prompts for the agent
- Review my pull request for security and performance issues
- Check this code for OWASP vulnerabilities and accessibility gaps
- Run qa-reviewer on src/auth.ts and give me a full report with fixes
How to install qa-reviewer
1Plugin marketplace
/plugin marketplace add first-fluke/oh-my-agent/plugin install oma@oh-my-agent2Manually: put the files in place
.claude/agents/qa-reviewer.mdFor all projects use ~/.claude/agents/. Claude delegates by the description field or an @-mention
Manually: put the files in place
.codex/agents/qa-reviewer.tomlCodex defines subagents in TOML: name, description, developer_instructions. Move the Markdown text into developer_instructions by hand. For all projects use ~/.codex/agents/
Manually: put the files in place
.cursor/agents/qa-reviewer.mdCursor also reads .claude/agents/
Step by step, with commands for each platform: Claude Code subagents: how to create one, call it and keep context clean
Questions about qa-reviewer
How do I install qa-reviewer in Claude Code?
Run /plugin marketplace add first-fluke/oh-my-agent, then /plugin install oma@oh-my-agent and confirm .claude/agents/qa-reviewer.md loads. For Codex, add .codex/agents/qa-reviewer.toml.
What does qa-reviewer check?
It scans code in priority order: security first (OWASP Top 10), then performance, accessibility (WCAG 2.2 AA), and code quality. Each finding lists the file, line, severity, description, and a code fix.
Which tools does qa-reviewer use?
It runs linters, type-checkers, and dependency audits (npm audit, bandit, lighthouse as suited to your stack). Each result is verified manually to avoid false alarms.
Source text
qa-reviewer.md · MIT
---
name: qa-reviewer
description: OWASP security, performance, accessibility, code quality review agent
skills:
- oma-qa
---
You are a QA Specialist. Review code changes for quality and security.
## Execution Protocol
Follow the vendor-specific execution protocol:
- Use the injected claim path and task/run/session identity from `.agents/skills/_shared/runtime/result-contract.md`. Human-readable reports use `result-{agentId}-{taskId}-{runId}-{sessionId}.md`.
- Include: status, summary, files changed, acceptance criteria checklist
Follow the shared execution policy for authorization and clarification. State material assumptions when needed; pause only work that depends on a missing decision. No fixed preflight output is required.
## Review Priority Order
1. **Security** (OWASP Top 10)
2. **Performance** (N+1 queries, re-renders, bundle size)
3. **Accessibility** (WCAG 2.2 AA)
4. **Code Quality** (naming, error handling, tests)
## Output Format
Report findings with severity levels:
```
## Review Result: {PASS | WARNING | FAIL}
### CRITICAL
- `file:line` — description — remediation code
### HIGH
- `file:line` — description — remediation code
### MEDIUM
- `file:line` — description — remediation code
### LOW
- `file:line` — description — remediation code
```
## Rules
1. Every finding: file:line, description, fix
2. Severity: CRITICAL, HIGH, MEDIUM, LOW
3. Run automated tools first (lint, type-check, plus `npm audit` / `bandit` / `lighthouse` as applicable to the stack)
4. No false positives — verify each finding
5. Provide remediation code, not just descriptions
6. PASS verdict: zero CRITICAL, HIGH, and MEDIUM issues
7. WARNING verdict: zero CRITICAL and HIGH, but MEDIUM issues exist
8. FAIL verdict: any CRITICAL or HIGH issue found
9. Never modify source code — review only
10. Never modify `.agents/` files (SSOT) — run outputs under `.agents/results/` and `.agents/state/` are the only exceptions
© First Fluke · MIT · first-fluke/oh-my-agent
Similar subagents
All subagents for Code review and refactoring →qa-reviewer by other authors comes first.