feat(adapter): add LangChain export and run adapter

[dahlinomine]

Summary

Adds langchain as a new export format and run adapter.

Closes #2.

Changes

• src/adapters/langchain.ts — exports a gitagent directory as a LangChain create_tool_calling_agent + AgentExecutor Python script
• src/runners/langchain.ts — runs the exported script via python3
• Skills become @tool-decorated stub functions
• Compliance constraints injected into the ChatPromptTemplate system message
• Model routing: gpt-4o default, claude-* → ChatAnthropic, gemini-* → ChatGoogleGenerativeAI
• Also wires langgraph into adapters/index.ts, commands/export.ts, commands/run.ts (both adapters share the same wiring PR since they have overlapping index changes)

Usage

# Export to a LangChain Python script
gitagent export -f langchain -d ./my-agent -o agent.py

# Run directly
gitagent run -a langchain -d ./my-agent

Generated output example

from langchain.agents import create_tool_calling_agent, AgentExecutor
from langchain_core.prompts import ChatPromptTemplate, MessagesPlaceholder
from langchain_core.tools import tool
from langchain_openai import ChatOpenAI

@tool
def research(input: str) -> str:
    """Research a topic..."""
    raise NotImplementedError("research not yet implemented")

tools = [research]
llm = ChatOpenAI(model="gpt-4o", temperature=0.3)
prompt = ChatPromptTemplate.from_messages([...])
agent = create_tool_calling_agent(llm, tools, prompt)
agent_executor = AgentExecutor(agent=agent, tools=tools, verbose=True)

Testing

• npm run build passes
• gitagent export -f langchain -d examples/standard produces valid Python
• gitagent export -f langchain -d examples/full renders compliance constraints in system prompt

Checklist

• Follows the existing adapter pattern (openai, crewai, lyzr)
• Runner follows the existing runner pattern (runners/openai.ts)
• Wired into adapters/index.ts, commands/export.ts, commands/run.ts
• No new dependencies added
• Conventional commit message