Agentic QE Framework logo

Agentic QE Framework

Open Source

Open-source framework for building AI-powered quality engineering agents that author and run tests.

Visit websiteGitHub

Pricing

Free / Open source

Type

Automation

Languages

Python

// VERDICT

Reach for the Agentic QE Framework when you already work in an AI coding agent and want QE-specialised agents (test generation, coverage analysis, flaky detection) inside that workflow. Skip it when you don't use a coding agent or want a standalone/commercial tool.

Best for

An open-source package of specialised QE agents and skills that plug into your coding agent (works best with Claude Code) via MCP - generating tests, finding coverage gaps and detecting flaky tests from inside your AI dev workflow.

Avoid when

You don't use a coding agent, you want a standalone test runner, or you need a polished commercial platform.

CI/CD fit

MCP package for coding agents - not a standalone CI runner

Languages

Python

Team fit

Teams using AI coding agents · Claude Code users · QE engineers experimenting with agentic workflows

Setup

Medium

Maintenance

Low

Learning

Intermediate

Licence

Free / Open source

// BEST FOR

  • Adding QE-specialised agents to an existing AI coding agent via MCP
  • Generating tests (unit/integration/BDD) from inside your dev workflow
  • Finding coverage gaps with risk-weighted analysis
  • Detecting and analysing flaky tests with AI assistance
  • Open-source and free to fork, build and extend
  • Teams experimenting with agentic, autonomous QE workflows

// AVOID WHEN

  • You don't use an AI coding agent (it plugs into one)
  • You want a standalone test runner, not an agent package
  • A polished commercial platform with support is required
  • You want a GUI rather than a coding-agent/CLI workflow
  • Deterministic, non-AI tooling is mandated
  • You won't review agent-generated output critically

// QUICK START

# Install and initialise within your coding-agent workflow (MCP), e.g.:
npx agentic-qe init --auto
# then invoke QE agents from your coding agent (works best with Claude Code).

// ALTERNATIVES TO CONSIDER

ToolChoose it when
TestZeusYou want an open-source autonomous testing agent (Hercules) rather than a coding-agent plugin.
KushoAIYou want hosted AI generation of API test suites.
CodeRabbitYou want AI review of PRs rather than agentic test generation.

// FEATURES

  • Building blocks for plan/act/verify agent loops
  • Configurable LLM provider and tool adapters
  • Examples and starter agents for common QE tasks
  • Self-hosted execution with no managed-service dependency
  • Open source under a permissive license

// PROS

  • Full transparency — every prompt and tool call is inspectable
  • No vendor lock-in or per-seat pricing
  • Good fit for teams already running their own LLM infrastructure
  • Encourages experimentation with agent design patterns

// CONS

  • Early-stage project — small community and limited docs
  • Significant engineering effort to operationalise
  • No managed runtime, observability, or commercial support

// EXAMPLE QA WORKFLOW

  1. Install the package and connect it to your coding agent via MCP

  2. Invoke QE agents (test generation, coverage, flaky detection)

  3. Review the agent-generated tests and analyses critically

  4. Keep the meaningful output; discard noise

  5. Run generated tests in your normal CI

  6. Iterate as the open-source project evolves

// RELATED QA.CODES RESOURCES

Cheat sheets