🎼 UI + API test automation
Explores an application (frontend through Playwright MCP, backend through its OpenAPI spec), plans coverage, implements UI and API tests, reviews them and runs them to a verified result. Pick it for a full suite with a review and verification stage; for a small first batch of tests, or a project that still needs a test framework set up, use ui-api-automation-minimal.
Install: ui-api-automation@jaktestowac. See the quick start for each tool.
Start the pack
| Tool | How |
|---|---|
| GitHub Copilot | Pick QA Orchestrator in the agent dropdown and describe the job. It hands off to the other agents. |
| Claude Code | Run /ui-api-automation followed by the job. The command coordinates; the agents below do the work. |
| OpenAI Codex | Copy codex/agents/ into .codex/agents/, then run $ui-api-automation followed by the job. |
Claude Code and Codex agents cannot start other agents, so there the orchestrator is a command that runs in your conversation.
Agents
| Agent | Name in Claude Code and Codex | What it does |
|---|---|---|
| QA Orchestrator (orchestrator) | /ui-api-automation command | Coordinates exploration, planning, implementation, review and verification of UI and API tests across the pack's subagents, and returns one summary. Use when a web application with an OpenAPI spec needs a planned, reviewed and verified Playwright UI + API suite. |
| BE Test Implementer | uiapi-be-implementer / uiapi_be_implementer | Implements API tests for the BE tasks of a test plan using the project's API test stack, and runs them before handing back. Use as the backend implementation step of the UI + API pack, once a plan exists; UI tests go to the FE Test Implementer. |
| FE Explorer (Playwright MCP) | uiapi-fe-explorer / uiapi_fe_explorer | Explores the running frontend through Playwright MCP, maps pages and user journeys, and records locator and flakiness risks plus prioritized FE test cases in .ai-outputs/. Use as the frontend exploration step of the UI + API pack, before planning. |
| FE Test Implementer | uiapi-fe-implementer / uiapi_fe_implementer | Implements Playwright UI tests for the FE tasks of a test plan, following the repository's conventions, and runs them before handing back. Use as the frontend implementation step of the UI + API pack, once a plan exists; API tests go to the BE Test Implementer. |
| OpenAPI Explorer | uiapi-openapi-explorer / uiapi_openapi_explorer | Analyzes an OpenAPI/Swagger spec and the related code into an endpoint inventory and a prioritized BE test matrix, written to .ai-outputs/. Use as the backend exploration step of the UI + API pack, before planning. |
| Solution Reviewer | uiapi-solution-reviewer / uiapi_solution_reviewer | Reviews the UI and API tests the pack implemented for assertion quality, locator robustness, flakiness, data isolation, secrets and fit with the plan, and returns tiered findings without editing. Use as the review step of the UI + API pack, after implementation and before the final run. |
| Test Planner | uiapi-test-planner / uiapi_test_planner | Combines the OpenAPI and frontend exploration summaries into one prioritized, implementation-ready test plan with an FE/BE task split, written to .ai-outputs/. Use as the planning step of the UI + API pack, after both explorers have reported. |
| Test Runner & Verifier | uiapi-test-runner / uiapi_test_runner | Runs the new and affected test suites, diagnoses each failure from its output and trace, and reports pass/fail with suspected causes without editing code. Use as the final verification step of the UI + API pack. |
MCP servers
playwright:npx @playwright/mcp@latest
Claude Code and Codex start these from the plugin. In GitHub Copilot, the agents that need them configure them themselves.
How the pack works, step by step: GUIDE.md.