Best Testing Skills for Codex CLI — QA and Test Generation
The best SKILL.md skills for testing and QA in OpenAI's Codex CLI.
Best Testing Skills for Codex CLI — QA and Test Generation
The best SKILL.md skills for testing and QA in OpenAI's Codex CLI.
Quick Answer: Best testing skills for Codex CLI: code-reviewer for automated review (447 installs), lobster-debugging for finding root causes (6 installs), ai-automation-qa-pack for QA documentation ($5), and truth-first for verification-driven testing (11 installs).
Codex CLI is OpenAI's terminal-based coding agent. It reads SKILL.md files from ~/.codex/skills/ (personal) and .codex/skills/ (project). Testing is one of the strongest use cases for skills in Codex CLI because the model writes better tests when given explicit patterns and conventions to follow.
Test review and code quality
code-reviewer (447 installs) reviews code for bugs, security vulnerabilities, and style issues. Use it to review tests themselves: it catches assertions that don't test anything meaningful, tests coupled to implementation details, and tests that would pass even with broken code. Free.
codex-grade-coding (37 installs) enforces strict verification protocols. The agent tests its own output before declaring a task complete. Free.
Recommended skills

Webapp Tester
by JustHandled Labs
Run practical browser tests for Claude Code and Codex projects, catching broken…

windows-desk-automation
by Roy Yuen
Reliable UIA-based Windows desktop automation with OCR and image matching fallba…

shift-left
by Matthew King
Risk analysis reporting and software test prioritisation
Debugging and root cause analysis
lobster-debugging (6 installs) is a systematic 4-phase debugging framework. It finds root causes, eliminates flaky tests, and prevents regressions. The structured approach helps Codex CLI focus instead of guessing. Free.
truth-first (11 installs) forces evidence-first verification. Before the agent changes a test, it checks what the actual system state is. Prevents fixing symptoms instead of causes. Free.
env-doctor (86 installs) diagnoses environment issues that cause test failures. If your CI passes but local tests fail, this finds the discrepancy. Free.
QA documentation
ai-automation-qa-pack (6 installs) generates professional QA and UAT documentation. Useful for teams that need formal test plans. $5.
Security testing
security-first (8 installs) forces early security framing. In a testing context, this means the agent generates security-focused test cases (SQL injection, XSS, CSRF) alongside functional tests. Free.
skill-security-vendor-pack (10 installs) audits code for security risks with professional reports. Free.
Installing testing skills in Codex CLI
Place skills in ~/.codex/skills/ for personal use or .codex/skills/ in your project:
~/.codex/skills/
├── code-reviewer/
│ └── SKILL.md
├── lobster-debugging/
│ └── SKILL.md
For a full guide, read Best Codex CLI Skills in 2026.
Browse all testing skills on Agensi.
Skills you might need
uBrowser
$15Run compact browser automation workflows with batched actions, reusable sessions…
GuardrailDoctor
$29Automated security auditing and penetration testing for Claude Code agent guardr…
finalize
$5The ultimate pre-commit checklist agent for cleaning code, updating docs, and va…
Related Articles
Best DevOps Skills for Codex CLI — CI/CD, Docker, and Deployment
The best SKILL.md skills for DevOps and deployment in OpenAI's Codex CLI.
4 min read
Best AI Agent Skills for Solo Developers (2026)
The best SKILL.md skills for solo developers and indie hackers. Multiply your output without hiring.
5 min read