Your agentic application-level reliability engineer.
Explore Filibuster · See the agent-led walkthrough · Understand the workflow
Your agent writes the feature and the end-to-end test. Filibuster explores dependency failures and shows the assumptions behind the result: which calls changed, which effects are missing, and what the customer experiences.
You inspect the healthy and faulted executions beside the agent's explanation, then decide what the application should do. The agent repairs the implementation or records your conditional expectation in the test, replays the failure and retained cases, and explains how the requirements became more precise.
A new release is in preparation. The Go and TypeScript implementations are being qualified. Public packages and installation instructions are coming next; the website and walkthrough show the current experience and its limits.
Earlier Filibuster implementations are archived for reference. They preserve the original research and are separate from the release now in development.
