Structured output and tool invocation

[ZYJ-3721]

Checked other resources

• This is a feature request, not a bug report or usage question.
• I added a clear and descriptive title that summarizes the feature request.
• I used the GitHub search to find a similar feature request and didn't find it.
• I checked the LangChain documentation and API reference to see if this feature already exists.
• This is not related to the langchain-community package.

Package (Required)

• langchain
• langchain-openai
• langchain-anthropic
• langchain-classic
• langchain-core
• langchain-cli
• langchain-model-profiles
• langchain-tests
• langchain-text-splitters
• langchain-chroma
• langchain-deepseek
• langchain-exa
• langchain-fireworks
• langchain-groq
• langchain-huggingface
• langchain-mistralai
• langchain-nomic
• langchain-ollama
• langchain-perplexity
• langchain-prompty
• langchain-qdrant
• langchain-xai
• Other / not sure / general

Feature Description

Currently, there is a conflict between the structured output of create_agent and the tool call. Can it be optimized in this direction: 1. Place the restrictions on the structured output at the last output. 2. Set structured output for the tool. Within the agent, the llm can call different tools and use different structured outputs as the input for the tool functions. A bit of insight, I wonder if it's feasible.

Use Case

The tool I invoked requires strictly structured input, which llm cannot achieve when calling

Proposed Solution

Currently, there is a conflict between the structured output of create_agent and the tool call. Can it be optimized in this direction: 1. Place the restrictions on the structured output at the last output. 2. Set structured output for the tool. Within the agent, the llm can call different tools and use different structured outputs as the input for the tool functions. A bit of insight, I wonder if it's feasible.

Alternatives Considered

No response

Additional Context

No response

[keenborder786]

I am not so sure I understand your request but just to differentiate.
If you use ProviderStrategy then it will use Provider Specific Structured Output without using the Tool calls, as oppose to ToolStrategy in which the arguments to a tool calls are parsed and given as final output.