Cognio MCP Server

MCP (Model Context Protocol) server wrapper for Cognio semantic memory.

Quick Setup

Run the auto-setup script to configure all supported AI clients:

npm run setup

This automatically generates MCP configurations for:

• Claude Desktop
• Claude Code (CLI)
• Cursor
• Continue.dev
• Cline
• Windsurf
• Kiro
• VS Code (GitHub Copilot)
• Gemini CLI

Manual Configuration

If you prefer to configure manually, add Cognio to your client's MCP config:

VS Code (GitHub Copilot)

Add to .vscode/mcp.json:

{
  "servers": {
    "cognio": {
      "command": "npx",
      "args": ["-y", "@0xrelogic/cognio-mcp"],
      "env": {
        "COGNIO_API_URL": "http://localhost:8080",
        "COGNIO_API_KEY": "your-optional-api-key-here"
      }
    }
  }
}

Claude Desktop

Add to ~/Library/Application Support/Claude/claude_desktop_config.json (Mac) or %APPDATA%/Claude/claude_desktop_config.json (Windows):

{
  "mcpServers": {
    "cognio": {
      "command": "npx",
      "args": ["-y", "@0xrelogic/cognio-mcp"],
      "env": {
        "COGNIO_API_URL": "http://localhost:8080"
      }
    }
  }
}

Cursor

Add to ~/.cursor/mcp_settings.json:

{
  "mcpServers": {
    "cognio": {
      "command": "npx",
      "args": ["-y", "@0xrelogic/cognio-mcp"],
      "env": {
        "COGNIO_API_URL": "http://localhost:8080"
      }
    }
  }
}

Continue.dev

Add to ~/.continue/config.json:

{
  "mcp": [
    {
      "name": "cognio",
      "command": "npx",
      "args": ["-y", "@0xrelogic/cognio-mcp"],
      "env": {
        "COGNIO_API_URL": "http://localhost:8080"
      }
    }
  ]
}

Cline

Add to ~/.cline/mcp.json:

{
  "mcpServers": {
    "cognio": {
      "command": "npx",
      "args": ["-y", "@0xrelogic/cognio-mcp"],
      "env": {
        "COGNIO_API_URL": "http://localhost:8080"
      }
    }
  }
}

Windsurf

Add to ~/.windsurf/mcp_config.json:

{
  "mcpServers": {
    "cognio": {
      "command": "npx",
      "args": ["-y", "@0xrelogic/cognio-mcp"],
      "env": {
        "COGNIO_API_URL": "http://localhost:8080"
      }
    }
  }
}

Kiro

Add to ~/.kiro/settings/mcp.json:

{
  "mcpServers": {
    "cognio": {
      "command": "npx",
      "args": ["-y", "@0xrelogic/cognio-mcp"],
      "env": {
        "COGNIO_API_URL": "http://localhost:8080"
      }
    }
  }
}

Gemini CLI

Add to ~/gemini/mcp.json:

{
  "mcpServers": {
    "cognio": {
      "command": "npx",
      "args": ["-y", "@0xrelogic/cognio-mcp"],
      "env": {
        "COGNIO_API_URL": "http://localhost:8080"
      }
    }
  }
}

Claude Code (CLI)

Add to ~/.claude.json:

{
  "mcpServers": {
    "cognio": {
      "type": "stdio",
      "command": "npx",
      "args": ["-y", "@0xrelogic/cognio-mcp"],
      "env": {
        "COGNIO_API_URL": "http://localhost:8080"
      }
    }
  }
}

Note: Claude Code requires "type": "stdio" in the config.

Available Tools

save_memory

Save information to long-term semantic memory with automatic tagging and categorization.

Parameters:

• text (required): The memory content to save
• project (optional): Project name for organization. Either provide this or use set_active_project first
• tags (optional): Array of tags. If omitted and auto-tagging is enabled with a valid LLM API key (GROQ_API_KEY or OPENAI_API_KEY), tags will be auto-generated
• metadata (optional): Key-value metadata object

Notes:

• A project is required (either via parameter or active project context)
• Auto-tagging requires: AUTOTAG_ENABLED=true and a configured LLM API key in .env
• If auto-tagging is disabled or misconfigured, memory saves without tags

search_memory

Search memories using semantic similarity.

Parameters:

• query (required): Search query text
• project (optional): Filter by project. If omitted, uses active project (required)
• tags (optional): Filter by tags array
• limit (optional): Max results (default: 5)

Notes:

• A project context is required (either via parameter or set_active_project)
• Similarity threshold is configurable via SIMILARITY_THRESHOLD in .env (default: 0.4)

list_memories

List all memories with optional filtering.

Parameters:

• project (optional): Filter by project. If omitted, uses active project (required)
• tags (optional): Filter by tags array
• limit (optional): Max results (default: 20)
• offset (optional): Skip results (default: 0)

Notes:

• A project context is required (either via parameter or set_active_project)

get_memory

Get a single memory by ID to view its full content.

Parameters:

• memory_id (required): The ID of the memory to retrieve

Notes:

• Use this when you need to read the complete text of a specific memory
• Memory IDs can be obtained from list_memories or search_memory results

get_memory_stats

Get statistics about stored memories.

No parameters required.

archive_memory

Archive (soft delete) a memory by ID.

Parameters:

• memory_id (required): The memory ID to archive

delete_memory

Permanently delete a memory by ID.

Parameters:

• memory_id (required): The memory ID to delete

export_memories

Export memories to JSON or Markdown format.

Parameters:

• format (optional): Export format - 'json' or 'markdown' (default: json)
• project (optional): Filter by project name

summarize_text

Summarize long text using extractive or abstractive methods.

Parameters:

• text (required): The text to summarize
• num_sentences (optional): Number of sentences in summary (default: 3, max: 10)

set_active_project

Set the active project context for all subsequent operations.

Parameters:

• project (required): Project name to activate

get_active_project

Get the currently active project context.

No parameters required.

list_projects

List all available projects in the database.

No parameters required.

Environment Variables

• COGNIO_API_URL: Base URL for Cognio API (default: http://localhost:8080)
• COGNIO_API_KEY: Optional API key for authenticated requests (sent as X-API-Key header if provided)

Configuration

Auto-tagging and other features are configured via .env in the Cognio project root:

# Auto-tagging (requires LLM API key)
AUTOTAG_ENABLED=true
LLM_PROVIDER=groq
GROQ_API_KEY=your-key-here
GROQ_MODEL=openai/gpt-oss-120b

# Or use OpenAI instead
# OPENAI_API_KEY=your-key-here
# OPENAI_MODEL=gpt-3.5-turbo

# Semantic search threshold (lower = more results, default 0.4)
SIMILARITY_THRESHOLD=0.4

# Summarization
SUMMARIZATION_ENABLED=true
SUMMARIZATION_METHOD=abstractive

See .env.example in the Cognio root directory for complete configuration options.

Requirements

• Node.js >= 18
• Cognio server running (default: http://localhost:8080)