langchain-ai/langgraph
1
0
Fork 0
mirror of https://github.com/langchain-ai/langgraph.git synced 2026-07-15 12:35:43 -04:00
Code Issues 314 Packages Projects Releases 17 Wiki Activity

[PR #5261] add Pydantic model support for ChatPromptTemplate in create_react_agent #4336

New Issue
Closed
opened 2026-02-20 17:50:02 -05:00 by yindo · 0 comments
yindo commented 2026-02-20 17:50:02 -05:00
Owner
Copy Link
Copy Source

Original Pull Request: https://github.com/langchain-ai/langgraph/pull/5261

State: closed
Merged: No

Problem

ChatPromptTemplate does not work properly with Pydantic models when used in prebuilt agents. The issue occurs because ChatPromptTemplate expects dictionary-like inputs for variable substitution (e.g., {role} placeholders), but Pydantic models are objects with attributes rather than dictionary keys.

Root Cause

When ChatPromptTemplate processes template variables like {role}, it tries to access them as dictionary keys. However, Pydantic models store data as object attributes, making them incompatible with the template's key-based variable substitution mechanism.

Implementation Details

File: libs/prebuilt/langgraph/prebuilt/chat_agent_executor.py

  • Added RunnableLambda import
  • Added conditional logic to detect Pydantic + BasePromptTemplate combination
  • Inserts conversion step in the model_runnable chain: pydantic_to_dict | prompt | model

File: libs/prebuilt/tests/test_react_agent.py

  • Added test_pydantic_prompt_and_state() to verify the fix works
  • Tests Pydantic state with ChatPromptTemplate variable substitution
  • Fixed minor syntax issue: changed return to pass in existing test

Testing

The new test case test_pydantic_prompt_and_state() verifies:

  • Pydantic state schema works with ChatPromptTemplate
  • Template variables are properly substituted from Pydantic model attributes
  • Expected output matches the templated result

Test scenario:

class State(BaseModel):
    messages: Annotated[list[AnyMessage], add_messages]
    role: str
    remaining_steps: RemainingSteps = 25

prompt = ChatPromptTemplate.from_messages([
    ("system", "You are a helpful assistant that knows the user's role is {role}"),
    ("placeholder", "{messages}"),
])

Backward Compatibility

This change is fully backward compatible:

  • Only activates when specific conditions are met (Pydantic + BasePromptTemplate)
  • TypedDict schemas continue to work unchanged
  • Non-Pydantic inputs are passed through without modification

Performance Impact

Minimal performance overhead:

  • Conversion only occurs when both conditions are met
  • isinstance() check is fast
  • model_dump() is efficient for Pydantic models

Related Issues

Closes #5230
Related to #3867

**Original Pull Request:** https://github.com/langchain-ai/langgraph/pull/5261 **State:** closed **Merged:** No --- ## Problem ChatPromptTemplate does not work properly with Pydantic models when used in prebuilt agents. The issue occurs because ChatPromptTemplate expects dictionary-like inputs for variable substitution (e.g., `{role}` placeholders), but Pydantic models are objects with attributes rather than dictionary keys. ## Root Cause When ChatPromptTemplate processes template variables like `{role}`, it tries to access them as dictionary keys. However, Pydantic models store data as object attributes, making them incompatible with the template's key-based variable substitution mechanism. ## Implementation Details **File: `libs/prebuilt/langgraph/prebuilt/chat_agent_executor.py`** - Added RunnableLambda import - Added conditional logic to detect Pydantic + BasePromptTemplate combination - Inserts conversion step in the model_runnable chain: `pydantic_to_dict | prompt | model` **File: `libs/prebuilt/tests/test_react_agent.py`** - Added `test_pydantic_prompt_and_state()` to verify the fix works - Tests Pydantic state with ChatPromptTemplate variable substitution - Fixed minor syntax issue: changed `return` to `pass` in existing test ## Testing The new test case `test_pydantic_prompt_and_state()` verifies: - Pydantic state schema works with ChatPromptTemplate - Template variables are properly substituted from Pydantic model attributes - Expected output matches the templated result Test scenario: ```python class State(BaseModel): messages: Annotated[list[AnyMessage], add_messages] role: str remaining_steps: RemainingSteps = 25 prompt = ChatPromptTemplate.from_messages([ ("system", "You are a helpful assistant that knows the user's role is {role}"), ("placeholder", "{messages}"), ]) ``` ## Backward Compatibility This change is fully backward compatible: - Only activates when specific conditions are met (Pydantic + BasePromptTemplate) - TypedDict schemas continue to work unchanged - Non-Pydantic inputs are passed through without modification ## Performance Impact Minimal performance overhead: - Conversion only occurs when both conditions are met - `isinstance()` check is fast - `model_dump()` is efficient for Pydantic models ## Related Issues Closes #5230 Related to #3867
yindo added the pull-request label 2026-02-20 17:50:02 -05:00
yindo closed this issue 2026-02-20 17:50:02 -05:00
Sign in to join this conversation.
No Branch/Tag Specified
Branches Tags
Labels
Clear labels
agents
awaiting author revision
bug
change
chat models
checkpointer
ci
code quality
code quality
dependencies
documentation
drawing
duplicate
enhancement
good first issue
help wanted
invalid
maintainer
meta
missing information
not maintained by langchain
open-swe-auto
open-swe-max
open-swe-max-auto
pending
performance
pull-request
Mirrored from GitHub Pull Request
 question
requires design
sdk
server
stale
studio
tests
type checking
usage question
v0.5
v1.1
No Label pull-request
Milestone
No items
No Milestone
Projects
Clear projects
No project
Assignees
Clear assignees
yindo
No Assignees
1 Participants
Notifications
Due Date
No due date set.
Dependencies

No dependencies set.

Reference: langchain-ai/langgraph#4336
Delete Branch "%!s()"

Deleting a branch is permanent. Although the deleted branch may continue to exist for a short time before it actually gets removed, it CANNOT be undone in most cases. Continue?