Tribal
by agentience
Tribal is an MCP server implementation for error knowledge tracking and retrieval. It provides both REST API and native MCP interfaces for integration with tools like Claude Code and Cline.
Last updated: N/A
What is Tribal?
Tribal is a Model Context Protocol (MCP) server designed to help Claude remember and learn from programming errors. It provides a knowledge service for storing, retrieving, and searching error records with full context, enhancing Claude's ability to solve coding problems.
How to use Tribal?
Tribal can be installed using uv package manager. It integrates with Claude via MCP, allowing Claude to store programming errors and solutions, search for similar errors, and build a knowledge base specific to your coding patterns. It can be run using the tribal command or directly via Python modules. API endpoints are available for creating, retrieving, updating, deleting, and searching errors. A client is also available to interact with the server.
Key features of Tribal
Store and retrieve error records with full context
Vector similarity search using ChromaDB
REST API (FastAPI) and native MCP interfaces
JWT authentication with API keys
Local storage (ChromaDB) and AWS integration
Docker-compose deployment
CLI client integration
Use cases of Tribal
Allow Claude to remember and learn from programming errors
Enable Claude to search for similar errors when encountering problems
Build a knowledge base specific to coding patterns
Integrate error tracking and retrieval into development workflows
FAQ from Tribal
How do I install Tribal?
How do I install Tribal?
Use the uv package manager to install directly from the project directory or build a distributable package.
How does Tribal integrate with Claude?
How does Tribal integrate with Claude?
Tribal integrates with Claude through the Model Context Protocol (MCP). Use the claude mcp add command to add Tribal.
What are the available MCP tools?
What are the available MCP tools?
Tribal provides tools like add_error, get_error, update_error, delete_error, search_errors, find_similar, and get_token.
Where are error records stored?
Where are error records stored?
Error records are stored in ChromaDB by default, with options for AWS S3 and DynamoDB integration.
How do I run the Tribal server?
How do I run the Tribal server?
You can run the server using the tribal command, Python modules, or legacy entry points like mcp-server and mcp-api.