Suitest - Self-Hosted Tests with Video Evidence

Open-source QA platform where agents generate, run, and fix tests with video evidence.

Unclaimed

Updated Jul 2026 · Added Jul 2026

suitest.suiflex.dev

Ai Test GenFreeai-test-gen

Visit Suitest

Share: X in LinkedIn

Add your screenshot here

Image or video shown in this spot

What is Suitest?

Suitest is an open-source, MCP-native QA platform that generates, runs, and publishes end-to-end tests with video evidence. It integrates directly with IDE agents like Claude Code and Cursor to automate test creation and execution against live applications. The platform runs fully self-hosted on your laptop with one command, scaling to Docker or Helm, and provides deterministic, repeatable testing with failure context serialized for AI-driven code fixes.

Explore Suitest

Need help implementing Suitest - Self-Hosted Tests with Video Evidence?

Find verified specialists who work with Suitest - Self-Hosted Tests with Video Evidence

Browse specialists

Key Features of Suitest

MCP-native IDE integration (Claude Code, Cursor, Codex)
One-command local setup with dashboard and SQLite
Automated end-to-end test generation from PRD
Deterministic browser automation with video recording
Failure context serialization for agent-driven fixes
Full local platform without cloud or LLM keys required
Docker Compose and Helm deployment options
Console error and network request capture
DOM excerpt extraction around failed selectors
Published test runs with evidence and history

Who Should Use Suitest?

Generate test suites from product requirements documents

Execute end-to-end tests against web applications

Capture and debug API failures during test runs

Automate selector validation and DOM verification

Publish QA evidence and test reports to team dashboard

Enable AI agents to identify and fix code bugs from test failures

CI/CD pipeline integration with deterministic replay

Suitest: Pros & Cons

✓Pros

Fully self-hosted with no external dependencies or cloud lock-in
No LLM key required to run locally
One-command setup and scaling
Deterministic and repeatable test execution
Rich failure context (video, screenshot, console, network, DOM)
Seamless IDE agent integration via MCP
Open-source transparency
Low token budget for failure serialization (under 8 KB)

✕Cons

Requires Node 18+ and uv for local setup
No cloud-hosted option mentioned for teams without infrastructure
Limited to self-hosted browser automation (potential scaling complexity)

Frequently Asked Questions about Suitest

What is an MCP server and how does Suitest use it?

MCP (Model Context Protocol) is how Suitest integrates with IDE agents like Claude Code, Cursor, and Codex. The MCP server allows your agent to analyze your app, generate tests, run them, and publish results without leaving the editor.

Do I need an LLM key to run Suitest?

No. Suitest runs fully locally on your laptop with Node 18+ and uv. You only need an LLM key if you want to use an agent in your IDE to generate tests.

How does Suitest help agents fix bugs?

When a test fails, Suitest captures rich context (failed step, DOM excerpt, console errors, network requests, video, screenshot) in a small payload (under 8 KB). Your agent reads this context and can directly edit the app code to fix the bug, then re-run the test.

Can Suitest scale beyond my laptop?

Yes. Start with one command locally, then deploy to Docker Compose or Helm when your team grows. The full platform runs the same way in all environments.

What file format should I use for test requirements?

Suitest generates test suites from a PRD.md file. You provide your product requirements in Markdown format, and Suitest generates end-to-end test cases automatically.