Jinni

v1.0.0Coding Agentsstable

Bring your project into LLM context - tool and MCP server

jinnimcpai-integration
Share:
271
Stars
0
Downloads
0
Weekly
0/5

What is Jinni?

Jinni is a Model Context Protocol (MCP) server that allows AI assistants like Claude, Cursor, and VS Code to bring your project into llm context - tool and mcp server

Bring your project into LLM context - tool and MCP server

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

Features

  • Bring your project into LLM context - tool and MCP server

Use Cases

Bring your project into LLM context efficiently.
Use tool and MCP server for context management.
Optimize token usage by smart project summarization.
smat-dev

Maintainer

LicenseApache-2.0
Languagepython
Versionv1.0.0
UpdatedMay 15, 2026
Statushealthy
Maintenanceactive

Works with

ClaudeOpenAIwindowsmacoslinux

Installation

Manual Installation

npx jinni

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 Jinni

Jinni is an MCP server and CLI tool that consolidates your entire project's source files into a single, well-structured context dump for AI assistants. Rather than having an AI read files one-by-one, Jinni gives it a comprehensive view of all relevant code with path headers, allowing models to reason about the full project structure at once. It automatically excludes binary files, dotfiles, build artifacts, and node_modules, and supports fine-grained inclusion/exclusion control via .contextfiles — a gitignore-style configuration specific to Jinni.

Prerequisites

  • Python 3.10 or higher (for the jinni package)
  • uv or pip for installation
  • An MCP-compatible client such as Claude Desktop or Claude Code
  • A software project directory to provide as context
1

Install Jinni

Install jinni via pip or uv. The package provides both the jinni CLI tool and the jinni-server MCP entry point.

pip install jinni
2

Add Jinni to your MCP client config

Register the jinni MCP server using uvx which runs it in an isolated environment. No additional arguments are needed — the model will pass the project root when calling the tool.

{
  "mcpServers": {
    "jinni": {
      "command": "uvx",
      "args": ["jinni-server"]
    }
  }
}
3

Restart your MCP client

Restart Claude Desktop or reload your MCP client after saving the config. The jinni server exposes two tools: read_context and usage.

4

Optionally create a .contextfiles configuration

Place a .contextfiles file in your project root to control what gets included. Use gitignore-style patterns — lines starting with ! exclude matches. Rules in .contextfiles take priority over .gitignore.

# .contextfiles example
src/**/*.py
docs/**/*.md
!**/__pycache__/**
!**/node_modules/**
5

Ask your AI to read project context

Invoke the read_context tool by asking the AI to read context for your project directory. You can target specific modules or pass glob patterns to limit scope.

Jinni Examples

Client configuration

Claude Desktop config for Jinni using uvx to run jinni-server in an isolated environment.

{
  "mcpServers": {
    "jinni": {
      "command": "uvx",
      "args": ["jinni-server"]
    }
  }
}

Prompts to try

Example prompts that trigger the read_context tool to load project files.

- "Read context for the project at /Users/me/projects/myapp"
- "Load context for just the src and docs directories in this project"
- "What does this project do? Read its context first."
- "Read context for /Users/me/myapp with patterns matching only *.py files"
- "List the files in /Users/me/myapp without reading their content"

Troubleshooting Jinni

The context output is too large and exceeds the model's context window

Pass a size_limit_mb parameter to read_context, or use the targets parameter to specify only relevant subdirectories. You can also use list_only=true to see what files would be included before loading all content.

uvx command not found when starting the MCP server

Install uv first: curl -LsSf https://astral.sh/uv/install.sh | sh on macOS/Linux. Then ensure uv is on the PATH that Claude Desktop uses. Alternatively install jinni with pip and use the command 'jinni-server' directly.

Sensitive files like .env or credentials are included in context

Create a .contextfiles in your project root and add exclusion rules like !.env and !**/*.pem. Rules starting with ! negate (exclude) matches and take priority over .gitignore rules.

Frequently Asked Questions about Jinni

What is Jinni?

Jinni is a Model Context Protocol (MCP) server that bring your project into llm context - tool and mcp server It connects AI assistants to external tools and data sources through a standardized interface.

How do I install Jinni?

Follow the installation instructions on the Jinni GitHub repository. Clone the repo, install dependencies, and add the server config to your AI client.

Which AI clients work with Jinni?

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

Is Jinni free to use?

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

Browse More Coding Agents MCP Servers

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

Quick Config Preview

{ "mcpServers": { "jinni": { "command": "npx", "args": ["-y", "jinni"] } } }

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

Read the full setup guide →

Ready to use Jinni?

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