Diagnose and fix common LangChain errors and exceptions.
Use when encountering LangChain errors, debugging failures,
or troubleshooting integration issues.
Trigger with phrases like "langchain error", "langchain exception",
"debug langchain", "langchain not working", "langchain troubleshoot".
Installer avec Codex ou Claude Copiez ce prompt, collez-le dans Codex, Claude ou un autre assistant, puis laissez-le vérifier la page du skill et l'installer pour vous.
Une commande directe contourne le prompt de vérification. Examinez la source avant de l'exécuter.
Diagnose and fix common LangChain errors and exceptions.
Use when encountering LangChain errors, debugging failures,
or troubleshooting integration issues.
Trigger with phrases like "langchain error", "langchain exception",
"debug langchain", "langchain not working", "langchain troubleshoot".
allowed-tools
Read, Write, Edit, Grep
version
1.0.0
license
MIT
author
Jeremy Longshore <jeremy@intentsolutions.io>
LangChain Common Errors
Overview
Quick reference for diagnosing and resolving the most common LangChain errors.
Prerequisites
LangChain installed and configured
Access to application logs
Understanding of your LangChain implementation
Error Reference
Authentication Errors
openai.AuthenticationError: Incorrect API key provided
# Cause: Invalid or missing API key
os
os.environ[] =
langchain_openai ChatOpenAI
llm = ChatOpenAI()
# Solution:
import
"OPENAI_API_KEY"
"sk-..."
# Set correct key
# Verify key is loaded
from
import
# Will raise error if key invalid
anthropic.AuthenticationError: Invalid x-api-key
# Cause: Anthropic API key not set or invalid# Solution:
os.environ["ANTHROPIC_API_KEY"] = "sk-ant-..."# Or pass directlyfrom langchain_anthropic import ChatAnthropic
llm = ChatAnthropic(api_key="sk-ant-...")
Import Errors
ModuleNotFoundError: No module named 'langchain_openai'
# Cause: Provider package not installed# Solution:
pip install langchain-openai
# For other providers:
pip install langchain-anthropic
pip install langchain-google-genai
pip install langchain-community
ImportError: cannot import name 'ChatOpenAI' from 'langchain'
# Cause: Using old import path (pre-0.2.0)# Old (deprecated):from langchain.chat_models import ChatOpenAI
# New (correct):from langchain_openai import ChatOpenAI
Rate Limiting
openai.RateLimitError: Rate limit reached
# Cause: Too many API requests# Solution: Implement retry with backofffrom langchain_openai import ChatOpenAI
from tenacity import retry, wait_exponential, stop_after_attempt
@retry(wait=wait_exponential(min=1, max=60), stop=stop_after_attempt(5))defcall_with_retry(llm, prompt):
return llm.invoke(prompt)
# Or use LangChain's built-in retry
llm = ChatOpenAI(max_retries=3)
# Cause: Pydantic model validation failed# Solution: Make fields optional or provide defaultsfrom pydantic import BaseModel, Field
from typing importOptionalclassOutput(BaseModel):
answer: str
confidence: Optional[float] = Field(default=None)
Chain Errors
ValueError: Missing required input keys
# Cause: Input dict missing required variables# Debug:
prompt = ChatPromptTemplate.from_template("Hello {name}, you are {age}")
print(prompt.input_variables) # ['name', 'age']# Solution: Provide all required keys
chain.invoke({"name": "Alice", "age": 30})
# Cause: Agent stuck in loop# Solution: Increase iterations or improve prompts
agent_executor = AgentExecutor(
agent=agent,
tools=tools,
max_iterations=20, # Increase from default 15
early_stopping_method="force"# Force stop after max
)
ToolException: Tool execution failed
# Cause: Tool raised an exception# Solution: Add error handling in tool@tooldefmy_tool(input: str) -> str:
"""Tool description."""try:
# Tool logicreturn result
except Exception as e:
returnf"Tool error: {str(e)}"
Memory Errors
KeyError: 'chat_history'
# Cause: Memory key mismatch# Solution: Ensure consistent key names
prompt = ChatPromptTemplate.from_messages([
MessagesPlaceholder(variable_name="chat_history"), # Match this
("human", "{input}")
])
# When invoking:
chain.invoke({
"input": "hello",
"chat_history": [] # Must match placeholder name
})
Debugging Tips
Enable Verbose Mode
import langchain
langchain.debug = True# Shows all chain steps# Or per-component
agent_executor = AgentExecutor(verbose=True)
Trace with LangSmith
# Set environment variables
os.environ["LANGCHAIN_TRACING_V2"] = "true"
os.environ["LANGCHAIN_API_KEY"] = "your-langsmith-key"
os.environ["LANGCHAIN_PROJECT"] = "my-project"# All chains automatically traced
Check Version Compatibility
pip show langchain langchain-core langchain-openai
# Ensure versions are compatible:# langchain >= 0.3.0# langchain-core >= 0.3.0# langchain-openai >= 0.2.0