Vigil
Every broken user session becomes a developer-ready bug report.
Vigil watches real user sessions, clusters repeated failures, and turns broken UX into a prioritised GitHub issue queue — before your users open a support ticket.
Visit VigilA collection of products and experiments from the lab. Some are live, some are open source, and all of them are built to make complex systems easier to understand.
Every broken user session becomes a developer-ready bug report.
Vigil watches real user sessions, clusters repeated failures, and turns broken UX into a prioritised GitHub issue queue — before your users open a support ticket.
Visit Vigil
Trace a failed agent run back to the decision that changed its outcome.
Agent-Causality records every run as a causal graph, so you can debug branching multi-agent systems by evidence instead of log order. It ships with a Python SDK, CLI, explorer, and an OpenCode capture plugin.
Visit Agent-Causality