Fetter provides a remote Model Context Protocol (MCP) server at https://mcp.fetter.io/mcp that gives AI coding agents real-time access to Python package vulnerability data. Built on fetter, it queries PyPI and OSV to surface known CVEs, CVSS scores, and safe versions so your agent can make informed dependency decisions as it writes code.
Real-time Python package and vulnerability data for AI coding agents.
2026-11-09
Sterne
—
Tools
—
Uptime
—
Tools:
most_recent_not_vulnerable: find the latest release of a package that is free of known vulnerabilities
is_vulnerable: check whether a specific pinned version has known CVEs
lookup: find available versions and their vulnerabilities for any package or specifier
Installation
The Fetter MCP server uses the HTTP transport and requires no local installation. Just register the remote URL with your MCP client.
Claude Code
claude mcp add --transport http fetter https://mcp.fetter.io/mcp
For any other MCP-compatible client, provide the following remote server URL using the HTTP transport:
https://mcp.fetter.io/mcp
Agent Usage
Once installed, the Fetter MCP tools are available to your AI agent during coding sessions. The agent can call them automatically when adding or auditing dependencies; no explicit tool invocation is required in your prompts.
Example prompts
"Add the latest safe version of requests to requirements.txt"
"Are there any known vulnerabilities in my current dependencies?"
"What is the most recent version of pillow with no CVEs?"
"Before pinning cryptography, check whether 42.0.5 is vulnerable"
The agent selects the appropriate tool based on context:
Adding a new package: most_recent_not_vulnerable to find a safe version
Validating a specific pinned version: is_vulnerable for a definitive answer
Auditing an existing specifier: lookup to see affected versions
most_recent_not_vulnerable
Find the most recent version of a package that has no known vulnerabilities. Provide only a package name and the server will search recent releases for a safe version. Useful when pinning a dependency to the latest clean release.
Parameters
package_name — package name only (no version specifier), e.g. "requests"
Check if a specific package version has known vulnerabilities. Requires an exact version specifier. Returns vulnerability IDs, summaries, CVSS scores, severity ratings, and reference URLs.
Parameters
dep_spec — exact version specifier, e.g. "requests==2.31.0"
Look up a package by name and optional version specifier to find which versions are available and whether they have known vulnerabilities. Supports specifiers such as "requests", "numpy>=2.0", or "flask==3.0.0".
Parameters
dep_specs — package name or version specifier
cvss_threshold — filter to vulnerabilities at or above this CVSS score (0–10)
max_observed_score — return only the highest CVSS score per version rather than all individual vulnerabilities
count — limit the number of recent versions checked
retain_passing — include versions with no known vulnerabilities in the results
Noch keine Tool-Liste. Remote-Server erhalten eine nach dem ersten Health-Check.
FAQ
Fragen
Wie installiere ich Fetter MCP?
Kopieren Sie auf dieser Seite die http-Konfiguration für Cursor, Claude Desktop, Claude Code oder VS Code. Sie nutzt die Installation des Herausgebers, keinen PlugMCP-Proxy.
Benötigt Fetter MCP einen API-Schlüssel?
In diesem Eintrag ist keine erforderliche Umgebungsvariable aufgeführt.
Welchen Transport verwendet Fetter MCP?
Fetter MCP wird über HTTP erreicht.
Wann wurde Fetter MCP zuletzt geprüft?
Zuletzt geprüft: 2026-11-09. Real-time Python package and vulnerability data for AI coding agents.