dependency-freshness-mcp
Dependency Freshness Checker is an MCP tool that tells AI coding agents whether an npm or PyPI package is outdated, providing the latest version, release dates, deprecation status, versions behind, and a dated change summary with citations. It solves the problem of LLMs being time-blind and recommending deprecated packages. It is for developers building AI agents or using MCP-compatible tools.
✨ Key features
- Checks npm and PyPI packages for outdated versions.
- Returns latest version, release date, and deprecation status.
- Provides versions behind and dated change summary with citations.
- MCP-native with one tool: check_dependency_freshness(packages).
- Reads only public registries and GitHub releases, no scraping.
- Pay-Per-Event pricing: $0.005 per package checked.
🎯 Use cases
- AI coding agents verifying package versions before writing code.
- Automated dependency audits in CI pipelines.
- Dashboards monitoring dependency freshness across projects.
- Assisting developers in upgrading dependencies with change summaries.
📦 Installation
🧰 Requirements: Node.js for local stdio MCP server; optional GitHub token for higher rate limits; no API keys required for hosted options.
To use as a local stdio MCP server, clone the repository and build it. Then configure your MCP client with the command pointing to the compiled entrypoint. For hosted options, no installation is needed.
🚀 Usage
{
"packages": [
{ "ecosystem": "npm", "name": "zod", "currentVersion": "3.22.0" },
{ "ecosystem": "pypi", "name": "fastapi" }
]
}
⚠️ Good to know
Currently supports only npm and PyPI ecosystems.
❓ FAQ
Does it scrape websites?
No, it reads only public registry APIs and GitHub releases, so there is no scraping or terms-of-service risk.
Which ecosystems are supported?
npm and PyPI are supported today.
Do I need a GitHub token?
No, it is optional and only raises the GitHub rate limit for heavier batches.
Can an agent call it directly?
Yes, that is the point. Use the check_dependency_freshness MCP tool via hosted MCP server or the Actor's own MCP endpoint.
📊 Repository
🤖 Overview, features, install steps and FAQ were generated from the project's README on Sep 4, 2026. Always check the original source before running commands.