agent-qa

About agent-qa

agent-qa is an open-source QA testing harness that lets developers write tests in natural language for web and mobile applications. Agents execute these tests by working from visible UI roles, labels, and screen state rather than fixed selectors. When a sub-action such as a click, fill, or select fails, the system re-observes the UI and attempts a different path within the same run, allowing tests to recover from UI drift and flaky interactions.

The tool builds execution memory from product, suite, and test observations across runs, and applies that context to future test executions, including memory curated from steps that were healed during a run. An action cache reuses validated plans across similar subsequent test runs to reduce planner work, token usage, and runtime overhead. It also supports running Node, Bun, Python, or Bash hooks inside sandboxed Docker containers for tasks like environment setup, API calls, fixture seeding, and state teardown.

agent-qa is intended for software teams and provides both a dashboard and CLI for developers, along with MCP and skills support for coding agents. Tests, configurations, hooks, memory, and suite logic are stored as version-controlled code. It supports OpenAI- and Anthropic-compatible endpoints, Gemini, local or open-source models, and subscription-based auth for Codex and Claude Code. Installation is via npm, with separate commands to install browser and mobile driver support.

Key features

• Write tests in natural language for web and mobile
• Self-healing test execution that re-observes UI on failed actions
• Builds execution memory from past runs to inform future tests
• Action cache reuses validated plans to reduce token usage and runtime
• Runs sandboxed Node, Bun, Python, or Bash hooks in Docker containers
• Bring your own LLM via OpenAI- and Anthropic-compatible endpoints, Gemini, or local models

Frequently asked questions