agent-triage is an open source tool for ranking and diagnosing failures in multi-agent systems. It provides a CLI, web dashboard, OTLP receiver, and optional LLM-based root-cause analysis, helping AI engineers quickly identify and prioritize issues.
agent-triage is a LLM eval & observability project. It focuses on diagnosing and prioritizing failures in multi-agent systems using trace analysis and LLM-based root-cause detection. It is built as an open-source project for AI engineers. agent-triage is open source under the MIT license. It runs on the command line and the web.
Behind agent-triage is thebharathkumar, and it first shipped in 2026. The project is developed in the open on GitHub with 48 commits in the last 90 days. Key capabilities include trace ranking, severity scoring, and web dashboard.
Summary written by a language model from the project’s public pages.
What PulseGate has recorded for this listing
Closest matches by what these projects do