An MCP server with a database that creates a project-oriented knowledge graph for AI assistants, preserving solutions, progress, and architecture with vector embeddings for semantic search and RAG capabilities.

Installation

UVX (Recommended)

uvx --from context-portal-mcp conport-mcp --mode stdio --workspace_id ${workspaceFolder}

From Source

git clone https://github.com/GreatScottyMac/context-portal.git
cd context-portal
uv venv
uv pip install -r requirements.txt

Development Mode

uv run --python .venv/bin/python --directory <path to context-portal repo> conport-mcp --mode stdio

Configuration

UVX Configuration (Recommended)

{
  "mcpServers": {
    "conport": {
      "command": "uvx",
      "args": [
        "--from",
        "context-portal-mcp",
        "conport-mcp",
        "--mode",
        "stdio",
        "--workspace_id",
        "${workspaceFolder}",
        "--log-file",
        "./logs/conport.log",
        "--log-level",
        "INFO"
      ]
    }
  }
}

Without workspace ID on startup

{
  "mcpServers": {
    "conport": {
      "command": "uvx",
      "args": [
        "--from",
        "context-portal-mcp",
        "conport-mcp",
        "--mode",
        "stdio",
        "--log-file",
        "./logs/conport.log",
        "--log-level",
        "INFO"
      ]
    }
  }
}

Development Configuration

{
  "mcpServers": {
    "conport": {
      "command": "uv",
      "args": [
        "run",
        "--python",
        ".venv/bin/python",
        "--directory",
        "<path to context-portal repo> ",
        "conport-mcp",
        "--mode",
        "stdio",
        "--log-file",
        "./logs/conport-dev.log",
        "--log-level",
        "DEBUG"
      ],
      "disabled": false
    }
  }
}

Features

Structured context storage using SQLite (one database per workspace)
Multi-workspace support via workspace_id
Vector data storage and semantic search capabilities
Backend for Retrieval Augmented Generation (RAG)
Project knowledge graph with explicit relationships
Database schema evolution using Alembic migrations
Automatic workspace detection
Custom project data storage (glossaries, specifications)
Solution and progress tracking
Prompt caching support for compatible LLM providers

Usage Examples

Initialize according to custom instructions

Create a projectBrief.md file to bootstrap initial project context

Notes

ConPort includes strategy files for various IDE environments (Roo Code, CLine, Windsurf Cascade) with specific custom instructions. The server automatically creates an SQLite database for each workspace and supports automatic workspace detection. The projectBrief.md file in the workspace root can be used to create initial project context.

context-portal MCP server