CocoSearch

Verified Integration

Author: @VioletCranberry•Category: ServerApplication

GitHub Repository

JSON-RPC 2.0

Protocol Standard

Sub-second

Execution Latency

Active

Operational Status

Tags:#Node.js #Docker #Git #Database #GitHub #Claude #Search #PostgreSQL #API #Python #Filesystem #AI

Client Configuration

— Connect CocoSearch to Claude Desktop or Cursor in seconds

{
  "mcpServers": {
    "cocosearch": {
      "command": "npx",
      "args": [
        "-y",
        "@modelcontextprotocol/server-cocosearch"
      ],
      "env": {}
    }
  }
}

Paste into ~/Library/Application Support/Claude/claude_desktop_config.json (macOS) or %APPDATA%\Claude\claude_desktop_config.json (Windows).

Architecture & Capabilities

System Overview

Indexes codebases into PostgreSQL using pgvector embeddings via Ollama to enable local-first hybrid semantic code search.

Indexed Date

9/16/2026

License

Open Source

Protocol Layer

stdio / SSE RPC

Frequently Asked Questions

Architecture and operational details for CocoSearch

CocoSearch is designed for privacy. All operations, including embedding generation with Ollama and data storage with PostgreSQL, run entirely on your local machine. There are no external API calls, cloud dependencies, or telemetry involved.

Related MCP Servers

Browse all servers →

@NBibikov

Cogvault

Manages fleet-grade local memory for AI agents using plain Markdown, offering hybrid recall, multi-tenancy, and fully local operation without cloud services.

ClaudeSearchDatabase+85 FAQs

Learn more

@Perseon-Labs

Collector Service

Collects data from edge devices.

ClaudeSearchDatabase+64 FAQs

Learn more

@mundurragacl

Connect

Enables AI assistants to interact with Amazon Connect contact centers via an MCP server, providing comprehensive tools for contact center management.

ClaudeSearchDatabase+75 FAQs

Learn more