LangWatch 是开源的 Agent 可观测性与评测平台:把每一次模型调用和工具调用串成完整链路追踪,再拿同一批用例做回归测试,看提示词或流程改动之后分数是涨还是掉。除了追踪与评测,还带路由与治理,团队里有多个 Agent 和编码助手时能在同一处看每个应用的表现。可自托管。 🔗 文档: https://docs.langwatch.ai | GitHub Stars: 5k+
LangWatch is an open-source observability and evaluation platform for agents: every model and tool call is stitched into a full trace, then the same test set is rerun after a prompt or workflow change to see whether scores went up or down. Beyond tracing and evals it adds routing and governance, so a team running several agents and coding assistants can watch each application in one place. Self-hostable. 🔗 Docs: https://docs.langwatch.ai | GitHub Stars: 5k+
LangWatch belongs to the Developer Tools category on hedirbase. Tagged with: Agent可观测性,LLM评测,开源,追踪.
LangWatch is an open-source observability and evaluation platform for agents: every model and tool call is stitched into a full trace, then the same test set is rerun after a prompt or workflow change