Overview
What is Kylas?
Kylas is a Model Context Protocol (MCP) server for Kylas CRM lead operations. It allows MCP clients like Cursor or Claude Desktop to create leads, search and filter leads, and look up users, products, and pipelines.
How to use Kylas?
Install with pip install -e . or from PyPI after publication. Set the KYLAS_API_KEY environment variable (required) and optionally KYLAS_BASE_URL. Run the server via python -m kylas_crm_mcp or use the provided Docker image. MCP clients communicate over stdio. Available commands include create_lead, search_leads, lookup_users, lookup_products, lookup_pipelines, get_pipeline_stages, and search_idle_leads.
Key features of Kylas
- Create leads with dynamic fields from user context
- Search and filter leads by multiple criteria
- Resolve user names to IDs for ownership fields
- Resolve product names and pipeline stages
- Find leads idle for N days
Use cases of Kylas
- Create new leads in Kylas CRM from an AI assistant
- Search for leads based on custom filters and criteria
- Automatically map user, product, and pipeline names to internal IDs
- Identify leads with no recent activity for follow-up
FAQ from Kylas
What are the system requirements for Kylas?
Python 3.10+ and a Kylas account with an API key.
How do I configure Kylas?
Set the KYLAS_API_KEY environment variable. Optionally set KYLAS_BASE_URL (default: https://api.kylas.io/v1). Configuration can also be placed in a .env file.
What transport does Kylas use?
Kylas uses stdio transport, the default for MCP. Clients start the server process and communicate via stdin/stdout.
Can Kylas be run with Docker?
Yes. Build the image with docker build -t kylas-crm-mcp . and run it, passing the API key via the -e flag: docker run -e KYLAS_API_KEY=your_key -i kylas-crm-mcp.
How do I publish Kylas to the MCP Marketplace?
Publish the package to PyPI, then use the mcp-publisher tool (login via GitHub and run mcp-publisher publish). Full instructions are in PUBLISHING.md.
