x1pm + LangChain
Shared workspace for LangChain agents
The problem
LangChain agents lack persistent, shared context. Each agent run starts fresh without access to team decisions, prior outputs, or coordinated task lists.
Setup
Up and running in under 2 minutes.
1
Get your API key from x1pm
Sign up at x1pm.com and generate an API key from your organization settings.
2
Add MCP tools to your LangChain agent
Use an HTTP MCP client to connect your agent to x1pm.
# Use an HTTP MCP client library for LangChain
# Connect to: https://x1pm.com/mcp
# With header: Authorization: Bearer YOUR_API_KEY
# Example with HTTP MCP tools
from langchain.agents import initialize_agent
from mcp_http_tools import MCPHTTPToolkit
toolkit = MCPHTTPToolkit(
url="https://x1pm.com/mcp",
headers={"Authorization": "Bearer YOUR_API_KEY"}
)
agent = initialize_agent(toolkit.get_tools(), llm, agent="zero-shot-react-description")3
Create your workspace files
Add context files your LangChain agents can read and write.
What you get
- LangChain agents read shared context before acting
- Outputs persist in structured markdown and CSV
- Multiple agents coordinate through the same workspace
- Works with any LangChain-compatible LLM
