---

name: ai-llm-agentic-tooling-langchain-langgraph description: Integrates LangChain/LangGraph for building LLM-powered agents and applications in Python, facilitating advanced logic and workflows. license: MIT compatibility: opencode metadata: version: "1.0.0" domain: coding triggers: langchain, langgraph, llm integration, agent, workflows, python archetypes:

• tactical
• generation anti_triggers:
• brainstorming
• vague ideation
• code golf
• over-engineering response_profile: verbosity: low directive_strength: high abstraction_level: operational role: implementation scope: implementation output-format: code related-skills: ai-llm-agentic-tooling-mcp

---

AI LLM Agentic Tooling with LangChain/LangGraph

Integrates LangChain and LangGraph to facilitate the development of LLM-powered agents and applications. This skill focuses on building advanced logic and workflows using these frameworks in Python.

Use Cases

Use this skill when:

• Creating complex workflows involving decision-making models.
• Building agents that require chaining multiple LLM calls efficiently.
• Integrating external tools or APIs within a conversational agent framework.

Implementation Patterns

This skill offers an integration guide for LangChain and LangGraph, enabling the development of LLM-powered agents in Python. It facilitates various advanced logic and workflows through examples and best practices.

Basic LangChain Agent Creation

This example demonstrates how to define a basic LangChain agent:

from langchain import LLMChain, PromptTemplate

# Define a simple LLM chain
prompt = PromptTemplate(input_variables=['input'], template="""You are a helpful assistant. Assist with: {input}.
""")

agent = LLMChain(prompt=prompt)

LangGraph Workflow Example

This section illustrates how to use LangGraph to define tasks:

from langgraph import Executor, Task

@Task
def fetch_data():
    return {'data_key': 'value'}

@Task
def process_data(data):
    return data['data_key'] + ' processed'

executor = Executor(tasks=[fetch_data, process_data])
executor.run()

Advanced Usage with Error Handling

Include retries and error logging in your agents. Use state management techniques to preserve data between tasks.

Constraints on Use

• Ensure prompt structures are maintained to maximize performance and clarity.
• Validate context objects to ensure they adhere to expected formats and types.

Metadata Updates

archetypes: tactical
anti_triggers:
  - vague conversation
  - overly generic request
response_profile:
  verbosity: medium
  directive_strength: high
  abstraction_level: operational

Basic LangChain Agent Creation

from langchain import LLMChain, PromptTemplate

# Define a simple LLM chain
prompt = PromptTemplate(input_variables=['input'], template="""You are a helpful assistant. Assist with: {input}.
""")

agent = LLMChain(prompt=prompt)

LangGraph Workflow Example

from langgraph import Executor, Task

# Define tasks and executor
@Task
def fetch_data():
    return {'data_key': 'value'}

@Task
def process_data(data):
    return data['data_key'] + ' processed'

executor = Executor(tasks=[fetch_data, process_data])
executor.run()

Advanced Usage with Error Handling

• Include retries and error logging in your agents.
• Use state management techniques to preserve data between tasks.