DooDooLamb News

Show HN: ResiliReplay - Chaos testing for AI agents and MCP servers

Brief published August 1, 2026 ยท Original source published July 31, 2026

Original reporting by ali110v at github.com.

Automated brief. Verify important details at the original source.

Show HN: ResiliReplay - Chaos testing for AI agents and MCP servers

What happened

ResiliReplay is an open-source TypeScript toolkit designed to chaos-test AI agents and MCP servers. It replays failures deterministically and converts broken execution traces into regression tests. The tool is described as model-agnostic and local-first, meaning it runs without requiring external services or cloud dependencies. It targets developers building or maintaining AI agent pipelines who need structured ways to reproduce and lock in failure scenarios rather than debugging them ad hoc.

Why it matters

Regression testing for AI agents remains an unsolved pain point for most teams. By turning live failure traces into deterministic, replayable tests, ResiliReplay offers a path toward treating agent reliability more like traditional software quality assurance. The MCP server support extends this to a protocol layer that is seeing rapid adoption across agent tooling.

What to watch

The published evidence cuts off mid-sentence, so the full feature set and any limitations around supported agent frameworks or MCP versions are unclear. Watch for documentation completeness and whether the project gains community traction as a standard chaos-testing baseline for agent developers.

Original source