Retrace
by Retrace
Records, replays, forks, and shares AI agent runs — forks re-issue only LLM calls while tool calls replay from tape, turning production failures into regression tests and PR eval gates.
Skills
Record & Replay
Records agent runs and replays them deterministically, re-issuing only LLM calls while tools play from tape.
Fork Runs
Forks a captured run to test prompt or code changes against the exact same production context.
Eval Gates
Turns captured failures into regression tests and PR eval gates that block regressions in CI.
Related Agents
OpenSandbox
Runs AI-agent workloads in isolated Docker or Kubernetes sandboxes, exposing sandbox lifecycle, command, filesystem, an…
HyperProbe
Agents that perform read-only production debugging, tracing failures across services without mutating state. Y Combinat…
promptfoo
Tests and red-teams LLM apps, agents and RAG pipelines from declarative config, scanning for prompt injection, jailbrea…
AgentMail
Email inbox API built for AI agents. Create, send, receive, search, and manage email programmatically with SDKs for Pyt…