test_driven_qa — Claude Code subagent: QA plans and setup
test_driven_qa — a subagent for Claude Code turns requirements and repository materials into a QA plan or test artifact. It ranks test scenarios and keeps observed evidence separate from proposed checks.
no license ★ 776 updated 2026-11-12
.claude/agents/test_driven_qa.mdWhat test_driven_qa does
The subagent examines acceptance criteria, relevant code, existing tests, configuration, and dependency details, then writes a Markdown QA report. The report covers risks, ranked scenarios, needed fixtures and mocks, environment assumptions, coverage gaps, and explicit pass/fail criteria. It can also give reproduction steps for a supplied failure.
Where relevant, its scenarios cover boundary conditions and failures, including empty or oversized input, timeouts, and upstream errors. The report depends on the requirements and repository context provided; it does not claim tests ran without execution evidence. Its response contains the requested artifact, with no greeting or hidden reasoning.
What it helps with
- draft a ranked test plan for a feature change
- identify boundary and failure cases from acceptance criteria
- define coverage gaps and pass/fail criteria
- write reproduction steps for a reported failure
Example prompts for the agent
- Create a QA plan from these acceptance criteria and the repository code.
- Review this test output and write steps to reproduce the failure.
- List boundary cases and the fixtures or mocks needed to test them.
How to install test_driven_qa
Manually: put the files in place
.claude/agents/test_driven_qa.mdFor all projects use ~/.claude/agents/. Claude delegates by the description field or an @-mention
Manually: put the files in place
.codex/agents/test_driven_qa.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/test_driven_qa.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 test_driven_qa
How do I install test_driven_qa in Claude Code?
Add the subagent file to your project at .claude/agents/test_driven_qa.md.
What does test_driven_qa do?
It produces a QA report with a risk assessment, ranked test scenarios, and pass/fail criteria. For a supplied failure, it also gives reproduction steps.
What information does test_driven_qa need?
Provide the requirements or change under test along with relevant repository context, such as code, tests, and configuration. Test output, coverage details, or failure traces can also be supplied.