Skip to content
LangChain Blog·· 2 h agoAI score58

LangSmith for Coding Agents tracks spend by pull request outcome

LangSmith for Coding Agents: Now track spend by outcome

AI summary

LangChain launched LangSmith for Coding Agents in public beta on LangSmith Cloud, which attributes each traced coding agent session's spend to the pull requests it contributed to. The dashboard aggregates cost per merged PR, breaks spend down by PR state and by work type, and flags sessions that never produced a PR as unattributed spend. It traces Claude Code, OpenAI Codex, Cursor, GitHub Copilot in VS Code and Deep Agents into one view, and setup requires installing a plugin per agent plus connecting GitHub.

Source:LangChain Blog · langchain.com