Context Portal

v1.0.0Knowledge & Memorystable

Context Portal (ConPort): A memory bank MCP server building a project-specific knowledge graph to supercharge AI assistants. Enables powerful Retrieval Augmented Generation (RAG) for context-aware development in your IDE.

context-portalmcpai-integration
Share:
763
Stars
0
Downloads
0
Weekly
0/5

What is Context Portal?

Context Portal is a Model Context Protocol (MCP) server that allows AI assistants like Claude, Cursor, and VS Code to context portal (conport): a memory bank mcp server building a project-specific knowledge graph to supercharge ai assistants. enables powerful retrieval augmented generation (rag) for context-aware dev...

Context Portal (ConPort): A memory bank MCP server building a project-specific knowledge graph to supercharge AI assistants. Enables powerful Retrieval Augmented Generation (RAG) for context-aware development in your IDE.

This server falls under the Knowledge & Memory category on MCPgee, the world's largest MCP server directory with 33,000+ servers.

Features

  • Context Portal (ConPort): A memory bank MCP server building

Use Cases

Project-specific knowledge graph building
RAG-powered development in IDE
GreatScottyMac

Maintainer

LicenseApache 2.0
Languagepython
Versionv1.0.0
UpdatedMay 20, 2026
Statushealthy
Maintenanceactive

Works with

ClaudeOpenAIwindowsmacoslinux

Installation

NPM

npx -y context-portal-mcp

Manual Installation

npx -y context-portal-mcp

Configuration

Configuration Details

Config File

claude_desktop_config.json

Performance

Response Metrics

Response Time< 200ms
ThroughputMedium

Resource Usage

Memory UsageLow
CPU UsageLow

How to Set Up and Use Context Portal

Context Portal (ConPort) is an MCP server that maintains a project-specific knowledge graph stored in a local SQLite database, giving AI assistants persistent memory across sessions within a single project. It exposes tools for storing and retrieving product context, active development state, architectural decisions with rationale, progress logs, system patterns, and arbitrary custom data — all linked together in a graph that supports full-text search. Developers use ConPort in their IDE to give coding assistants like Claude long-term project memory so they stop forgetting decisions and context between conversations.

Prerequisites

  • Python 3.10+ and uv installed (pip install uv) for the uvx execution method
  • An MCP client integrated into your IDE (e.g. Claude in VS Code, Cursor, or Windsurf)
  • A project directory with an absolute path — ConPort stores its SQLite database in the workspace root
  • Optional: a projectBrief.md file in your workspace root to seed the initial product context
1

Install uv if not already available

ConPort is best run with uvx, which is bundled with uv. Install uv using the official installer script.

curl -LsSf https://astral.sh/uv/install.sh | sh
2

Verify the package is accessible via uvx

Test that uvx can download and run the ConPort MCP server before adding it to your client config.

uvx --from context-portal-mcp conport-mcp --help
3

Add ConPort to your MCP client configuration

Edit your MCP client config file and add a ConPort entry, setting --workspace_id to the absolute path of your project directory.

{
  "mcpServers": {
    "conport": {
      "command": "uvx",
      "args": [
        "--from", "context-portal-mcp", "conport-mcp",
        "--mode", "stdio",
        "--workspace_id", "/absolute/path/to/your/project"
      ]
    }
  }
}
4

Create a projectBrief.md in your workspace

Add a Markdown file describing your project's purpose, tech stack, and goals. ConPort will automatically import this as the initial Product Context on first run.

5

Initialize ConPort in your AI session

Start a new conversation in your MCP client and ask it to initialize using your custom ConPort instructions. This triggers ConPort to load the project context and active state.

6

Log your first architectural decision

Use the log_decision tool to record an important decision with its rationale and implementation details, making it searchable for all future sessions.

Context Portal Examples

Client configuration

MCP client configuration for ConPort pointing to a specific project workspace.

{
  "mcpServers": {
    "conport": {
      "command": "uvx",
      "args": [
        "--from", "context-portal-mcp", "conport-mcp",
        "--mode", "stdio",
        "--workspace_id", "/Users/yourname/projects/my-app",
        "--log-level", "INFO"
      ]
    }
  }
}

Prompts to try

Example prompts to use ConPort's knowledge graph tools through your AI assistant.

- "Initialize according to custom instructions and show me the current project context"
- "Log the decision to use PostgreSQL instead of SQLite, with rationale about scalability"
- "What architectural decisions have we made about the authentication system?"
- "Search the project knowledge for any patterns related to error handling"
- "Show me a summary of recent activity and what we were working on last session"

Troubleshooting Context Portal

ConPort fails to start with 'workspace_id not found' error

The --workspace_id must be an absolute path to an existing directory. Relative paths are not supported. Use 'pwd' in your project directory to get the full path and paste it into the config.

uvx command not found when the MCP client starts the server

Install uv with 'curl -LsSf https://astral.sh/uv/install.sh | sh' and ensure ~/.cargo/bin or ~/.local/bin is in your PATH. Alternatively, use the full path to uvx in the 'command' field.

AI assistant does not use ConPort tools automatically

ConPort works best with custom system prompt instructions that tell the AI to load context at session start and save decisions/progress during the session. Add ConPort usage rules to your IDE's custom instructions or system prompt.

Frequently Asked Questions about Context Portal

What is Context Portal?

Context Portal is a Model Context Protocol (MCP) server that context portal (conport): a memory bank mcp server building a project-specific knowledge graph to supercharge ai assistants. enables powerful retrieval augmented generation (rag) for context-aware development in your ide. It connects AI assistants to external tools and data sources through a standardized interface.

How do I install Context Portal?

Install via npm with the command: npx -y context-portal-mcp. Then add the server configuration to your AI client's JSON config file (e.g., claude_desktop_config.json or .cursor/mcp.json).

Which AI clients work with Context Portal?

Context Portal works with all major MCP-compatible AI clients including Claude Desktop, Claude Code, Cursor, VS Code (GitHub Copilot), Windsurf, and Cline.

Is Context Portal free to use?

Yes, Context Portal is open source and available under the Apache 2.0 license. You can use it freely in both personal and commercial projects.

Browse More Knowledge & Memory MCP Servers

Explore all knowledge & memory servers available in the MCPgee directory. Each server includes setup guides for Claude, Cursor, and VS Code.

Quick Config Preview

{ "mcpServers": { "context-portal": { "command": "npx", "args": ["-y", "context-portal-mcp"] } } }

Add this to your claude_desktop_config.json or .cursor/mcp.json

Read the full setup guide →

Ready to use Context Portal?

Browse our complete directory of 33,000+ MCP servers, read setup guides for your editor, and start building with the Model Context Protocol.

33,000+ ServersFree & Open SourceStep-by-Step Guides