What is supported
LangChain 1.x chains andAgentExecutor (from langchain-classic): anything with invoke and input_keys. A LangChain agent made with create_agent is a LangGraph graph, and runs as one: see LangGraph.
A complete agent
agent.yml
requirements.txt
agent.py
OPENAI_API_KEY as a secret on the agent’s Config tab, then deploy. Any model works: this one is an example. See Configuration and secrets.
A complete LangChain example
expense-assistant: A tool calling agent run by an AgentExecutor, converting and recording expenses. It runs with no model key.The prompt and the reply
The prompt goes in the chain’sinput key, or its first input key if it has no input. A chain with chat_history gets an empty history.
Events it reports
See Streaming and events.
The caller’s fields
Underconfigurable.onecortex in the run’s config, for a chain that reads it. An AgentExecutor does not pass its config to tools, so an executor’s tools cannot read them: build the agent as a LangGraph graph instead.
Known limits
- The reply arrives once, at the end, not token by token.
- An
AgentExecutor’s tools cannot read the caller’s fields.
Troubleshooting
More on Troubleshooting and Errors.