Chroma MCP Server
About
An MCP server for the Chroma embedding database, providing persistent, searchable working memory for AI-assisted development with features like automated context recall and codebase indexing.
Details
- Author
- djm81
- GitHub stars
- 26
- Downloads
- 529
- Categories
- Database, Other, Knowledge Base, AI
Jump to
- Automated codebase indexing and change tracking
- Automated chat logging with enhanced context capture
- Bidirectional linking between discussions and code changes
- Semantic code chunking preserving logical code structures
- Working memory tools for capturing and retrieving context
- Validation system for evidence-based code change assessment
- Automated test-driven learning workflow from failure to fix
Setting up with Highlight
This MCP is not yet compatible with Highlight’s one-click setup. However, you can still use it with Highlight by following these steps:
- Download and install Highlight from highlightai.com/download
- Navigate to the plugins tab and select "Add Custom Plugin"
-
Configure the plugin with the settings below
Plugin Name
Chroma MCP ServerCommand (node, npx, python, etc.)Please refer to the README for specific instructions on how to obtain API keys or other required environment variables.
- Enable "Start Automatically" if you want the plugin to start when Highlight launches
From the repository
Install via pip install chroma-mcp-server (add [full] for all embedding models). Run with chroma-mcp-server --client-type ephemeral for in-memory storage or --client-type persistent --data-dir ./my_data for persistent storage. Integrate with Cursor by adding a configuration block to .cursor/mcp.json.
Claude Desktop / Cursor
Paste into your MCP client config file to install this server.
{
"mcpServers": {
"chroma mcp server": {
"chroma_mcp_prod": {
"command": "uvx",
"args": [
"chroma-mcp-server"
],
"env": {
"CHROMA_CLIENT_TYPE": "ephemeral",
"LOG_LEVEL": "INFO",
"MCP_LOG_LEVEL": "INFO"
}
}
}
}
}McpServers
{
"chroma_mcp_prod": {
"command": "uvx",
"args": [
"chroma-mcp-server"
],
"env": {
"CHROMA_CLIENT_TYPE": "ephemeral",
"LOG_LEVEL": "INFO",
"MCP_LOG_LEVEL": "INFO"
}
}
}A Model Context Protocol (MCP) server integration forChroma, the open-source embedding database.
Chroma MCP Server creates a persistent, searchable "working memory" for AI-assisted development:
- Automated Context Recall:AI assistants can query relevant information from past sessions
- Developer-Managed Persistence:Store key decisions and insights in ChromaDB via MCP
- Second Brain Integration:Integrates with IDE workflows to create a unified knowledge hub
- Automated Codebase Indexing:Track and index code changes
- Automated Chat Logging:Log AI interactions with enhanced context capture (code diffs, tool sequences)
- Bidirectional Linking:Connect discussions to code changes for tracing feature evolution
- Semantic Code Chunking:Preserve logical code structures for more meaningful context retrieval
- Working Memory Tools:MCP commands for capturing and retrieving development context
- Validation System:Evidence-based validation for code changes and learning promotions
- Automated Test-Driven Learning:Fully automated workflow from test failure to verified fix and learning promotion. See thePytest Plugin Usage Guideto integrate this into your projects.
See theGetting Started with your Second Brain guidefor more details.
# Basic installation pip install chroma-mcp-server # Full installation with all embedding models pip install "chroma-mcp-server[full]"
# With in-memory storage (data lost on restart) chroma-mcp-server --client-type ephemeral # With persistent storage chroma-mcp-server --client-type persistent --data-dir ./my_data
Add or modify.cursor/mcp.jsonin your project root:
{ "mcpServers": { "chroma": { "command": "uvx", "args": [ "chroma-mcp-server" ], "env": { "CHROMA_CLIENT_TYPE": "persistent", "CHROMA_DATA_DIR": "/path/to/your/data", "CHROMA_LOG_DIR": "/path/to/your/logs", "LOG_LEVEL": "INFO", "MCP_LOG_LEVEL": "INFO", "MCP_SERVER_LOG_LEVEL": "INFO" } } } }
- Enhanced Context Capture:Automatically extracts code diffs, tool sequences, and assigns confidence scores
- Bidirectional Linking:Creates navigable connections between chat discussions and code changes
- Semantic Code Chunking:Uses logical boundaries (functions, classes) instead of fixed-size chunks
- Server-Side Timestamp Enforcement:Ensures consistent timestamps across all collections
- Automatic Collection Creation:Essential collections (e.g.,chat_history_v1,codebase_v1) are automatically created on server startup if they don't exist.
- Enhanced Logging System:Per-execution log files prevent contamination of JSON communication in stdio mode
- Embedding Function Management:Tools to update collection metadata when changing embedding functions
- Collection Setup Command:Simplifies creation of multiple collections with consistent configuration
- Auto-Promote Workflow:Streamlined derived learning promotion with automatic handling of high-confidence entries
- Smart Defaults:Interactive promotion with intelligent defaults for all fields based on context
- Low Confidence Warnings:Visual indicators for entries that may need more careful review
- Automated Test Workflow:Fully automated capture of test failures, monitoring for fixes, and validated learning promotion
Comprehensive documentation is available in thedocs directory:
- Main Documentation- Complete guide to installation, configuration, and usage
- Getting Started- Detailed setup instructions
- Developer Guide- For contributors and developers
- IDE & Tool Integration Guides- Guides for integrating with IDEs and other tools.
- Automated Chat Logging- Enriched chat history with bidirectional linking
- Usage Guides- Detailed guides on how to use specific features and workflows.
- Enhanced Context Capture- Details on code diff extraction and tool sequencing
- Semantic Code Chunking- Logic-preserving code chunking for meaningful retrieval
- Automated Test Workflow (Pytest Plugin Usage)- Test-driven learning with automatic validation
- Thinking Tools & Utilities- Documentation for structured thinking and memory tools.
- Client and Developer Scripts- Guides for CLI tools and developer scripts.
- Logging Documentation- Overview of logging features and configuration.
- Server Logging- Details on the improved logging system
- Automation Documentation- Guides on automating development tasks.
- Project Rules & Guidelines- Development rules, guidelines, and best practices.
- Refactoring Plans- Documentation on various refactoring efforts and architectural plans.
- API Reference- Available MCP tools and parameters
Chroma MCP Server is licensed under the MIT License with Commons Clause. This means you can:
- Use Chroma MCP Server for any purpose (personal, commercial, academic)
- Modify the code
- Distribute copies
- Create and sell products built using Chroma MCP Server
- Sell Chroma MCP Server itself
- Offer Chroma MCP Server as a hosted service
- Create competing products based on Chroma MCP Server
See theLICENSE.mdfile for the complete license text.
Uses Qdrant vector embeddings to understand semantic relationships in codebases.
A production-ready Model Context Protocol (MCP) server that provides a powerful, vector-native memory bank for AI agents. Built with the Protocol-Lattice Go Agent Framework, this server offers persistent, searchable, and shareable memory with multiple database backends.
Fetch relevant embeddings and content from Supavec for AI assistants.
Embeddings, vector search, document storage, and full-text search with the open-source AI application database
Implement semantic memory layer on top of the Qdrant vector search engine
A vector database server powered by Chroma, enabling semantic document search, metadata filtering, and document management.
Provides AI assistants with persistent memory using ChromaDB vector storage.
An MCP server for vector storage and retrieval using ChromaDB.
Persistent memory layer for AI agents with semantic search, consolidation, and cross-session intelligence via MCP.
Sign in to leave a review
Use Google, GitHub, or an email account so ratings stay tied to real people.
No reviews posted yet.
