Developer releases agent-inspect, a local debugger for TypeScript AI agent traces

A developer has released agent-inspect, an open-source toolkit designed to help engineers debug and test TypeScript AI agent workflows locally. The tool addresses a common pain point where flat logs fail to capture the nested structure of multi-step agent pipelines, making it difficult to trace failures like timeouts or skipped retrievals. Agent-inspect converts a local JSONL trace file into a readable execution tree, a deterministic regression gate, and a shareable evidence bundle, all without requiring an account or uploading data by default. It supports popular frameworks including AI SDK, OpenAI Agents JS, LangChain, and LangGraph through adapters. The toolkit is available via npm and includes a synthetic demo that runs without making any model calls or network requests.
This is an AI-generated summary. ShortSingh links to the original source for the complete article.
Discussion (0)
Log in to join the discussion and vote.
Log in