LangSmith-Integration für KI-Agenten

AACFlow's LangSmith integration adds 5 tools to the AI agents you build in AACFlow's visual workflow builder. Forward workflow runs to LangSmith for observability.

You can create run, create runs batch, update run, get run, and create feedback with LangSmith in AACFlow. Each runs as a tool inside an AI agent, so the agent can combine LangSmith with any other connected service and apply LLM reasoning between steps.

(1) Create an account at aacflow.io. (2) Create an agent in your workspace. (3) Drag a LangSmith block onto the workflow builder. (4) Paste your LangSmith API key to authenticate. (5) Pick a tool such as "Create Run", wire up its inputs, and click Run.

Add a LangSmith block to your agent and select "Create Run" as the tool. Forward a single run to LangSmith for ingestion. Fill in the required fields. Inputs can reference outputs from earlier steps, such as text generated by an AI block or data fetched from another integration.