Fetter MCP
MCP serverSecurityFetter gives your AI real-time data about Python packages and their known vulnerabilities. Once added, your AI can pull current package and security information while it writes code, helping it catch risky dependencies early.
Use Fetter MCP in Claude, ChatGPT or Ahel Desktop
Free. Sign in, add Fetter MCP and connect your AI. About a minute.
Also: Claude Code · Cursor · Codex
Then ask your AI: use the most recent not vulnerable tool from Fetter MCP
No key or account is asked for. Not verified by us yet: Ahel has not run a call through it.
Details
Available today. Use it from your connected AI after setup.
No key or account is asked for. Not verified by us yet: Ahel has not run a call through it.
Add Fetter, then ask your AI to look up a Python package or check one for known vulnerabilities.
What your AI can do with it
- Look up current information about Python packages
- Check Python packages for known vulnerabilities
- Get up-to-date security data while writing code
- Spot packages with known issues before they are used
Install Fetter MCP
The server’s own address, for the clients that take one directly. Or connect Ahel once and every client you use reads it from one address, with the account kept on Ahel rather than in each client’s config.
Claude Code
claude mcp add --transport http --scope user fetter-mcp 'https://mcp.fetter.io/mcp'Run it once in your project, then open /mcp to approve any sign-in the server asks for.
Claude Desktop
https://mcp.fetter.io/mcpAdd a custom connector in Settings, paste this address, and approve the sign-in.
Cursor
cursor://anysphere.cursor-deeplink/mcp/install?name=fetter-mcp&config=eyJ1cmwiOiJodHRwczovL21jcC5mZXR0ZXIuaW8vbWNwIn0=Open the link and Cursor adds the server at that address.
ChatGPT
https://mcp.fetter.io/mcpIn Settings, enable Developer mode, create an MCP app, and paste this address. Your plan and workspace must allow custom apps.
Codex
codex mcp add fetter-mcp --url 'https://mcp.fetter.io/mcp'Run it once, then sign in with codex mcp login fetter-mcp if the server asks for an account.
From the project's README
As published by fetter-io/fetter-mcp in README.md.
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.
Tools:
most_recent_not_vulnerable: find the latest release of a package that is free of known vulnerabilitiesis_vulnerable: check whether a specific pinned version has known CVEslookup: 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
Codex
codex mcp add fetter --url https://mcp.fetter.io/mcp
Other MCP Clients
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_vulnerableto find a safe version - Validating a specific pinned version:
is_vulnerablefor a definitive answer - Auditing an existing specifier:
lookupto 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"
Example Request
{
"jsonrpc": "2.0",
"method": "tools/call",
"id": 2,
"params": {
"name": "most_recent_not_vulnerable",
"arguments": {
"name": "cryptography"
}
}
}
Example Response:
{
"jsonrpc": "2.0",
"id": 2,
"result": {
"content": [],
"structuredContent": {
"package": "cryptography",
"version": "46.0.5",
"vulnerabilities": [],
"vulnerable": false
},
"isError": false
}
}
is_vulnerable
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"
Example Request
{
"jsonrpc": "2.0",
"method": "tools/call",
"id": 2,
"params": {
"name": "is_vulnerable",
"arguments": {
"name": "requests==2.19.1"
}
}
}
Example Response:
{
"jsonrpc": "2.0",
"id": 2,
"result": {
"content": [],
"structuredContent": {
"package": "requests",
"version": "2.19.1",
"vulnerabilities": [
{
"cvss_score": 5.3,
"id": "GHSA-9hjg-9r4m-mvj7",
"severity": "(Medium):",
"summary": "Requests vulnerable to .netrc credentials leak via malicious URLs",
"url": "https://osv.dev/vulnerability/GHSA-9hjg-9r4m-mvj7"
},
{
"cvss_score": 5.6,
"id": "GHSA-9wx4-h78v-vm56",
"severity": "(Medium):",
"summary": "Requests Session object does not verify requests after making first request with verify=False",
"url": "https://osv.dev/vulnerability/GHSA-9wx4-h78v-vm56"
},
{
"cvss_score": 6.1,
"id": "GHSA-j8r2-6x86-q33q",
"severity": "(Medium):",
"summary": "Unintended leak of Proxy-Authorization header in requests",
"url": "https://osv.dev/vulnerability/GHSA-j8r2-6x86-q33q"
},
{
"cvss_score": 7.5,
"id": "GHSA-x84v-xcm2-53pg",
"severity": "(High):",
"summary": "Insufficiently Protected Credentials in Requests",
"url": "https://osv.dev/vulnerability/GHSA-x84v-xcm2-53pg"
},
{
"cvss_score": null,
"id": "PYSEC-2018-28",
"severity": null,
"summary": "",
"url": "https://osv.dev/vulnerability/PYSEC-2018-28"
},
{
"cvss_score": null,
"id": "PYSEC-2023-74",
"severity": null,
"summary": "",
"url": "https://osv.dev/vulnerability/PYSEC-2023-74"
}
],
"vulnerable": true
},
"isError": false
}
}
lookup
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 specifiercvss_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 vulnerabilitiescount— limit the number of recent versions checkedretain_passing— include versions with no known vulnerabilities in the results
Example Request
{
"jsonrpc": "2.0",
"method": "tools/call",
"id": 2,
"params": {
"name": "lookup",
"arguments": {
"name": "requests>=2.32.0",
"retain_passing": true
}
}
}
Example Response:
{
"jsonrpc": "2.0",
"id": 2,
"result": {
"content": [],
"structuredContent": {
"package": "requests",
"versions": [
{
"version": "2.32.0",
"vulnerabilities": [
{
"cvss_score": 5.3,
"id": "GHSA-9hjg-9r4m-mvj7",
"severity": "(Medium):",
"summary": "Requests vulnerable to .netrc credentials leak via malicious URLs",
"url": "https://osv.dev/vulnerability/GHSA-9hjg-9r4m-mvj7"
}
],
"vulnerable": true
},
{
"version": "2.32.1",
"vulnerabilities": [
{
"cvss_score": 5.3,
"id": "GHSA-9hjg-9r4m-mvj7",
"severity": "(Medium):",
"summary": "Requests vulnerable to .netrc credentials leak via malicious URLs",
"url": "https://osv.dev/vulnerability/GHSA-9hjg-9r4m-mvj7"
}
],
"vulnerable": true
},
{
"version": "2.32.2",
"vulnerabilities": [
{
"cvss_score": 5.3,
"id": "GHSA-9hjg-9r4m-mvj7",
"severity": "(Medium):",
"summary": "Requests vulnerable to .netrc credentials leak via malicious URLs",
"url": "https://osv.dev/vulnerability/GHSA-9hjg-9r4m-mvj7"
}
],
"vulnerable": true
},
{
"version": "2.32.3",
"vulnerabilities": [
{
"cvss_score": 5.3,
"id": "GHSA-9hjg-9r4m-mvj7",
"severity": "(Medium):",
"summary": "Requests vulnerable to .netrc credentials leak via malicious URLs",
"url": "https://osv.dev/vulnerability/GHSA-9hjg-9r4m-mvj7"
}
],
"vulnerable": true
},
{
"version": "2.32.4",
"vulnerabilities": [],
"vulnerable": false
},
{
"version": "2.32.5",
"vulnerabilities": [],
"vulnerable": false
}
]
},
"isError": false
}
}
Tools it offers (3)
What this server listed when Ahel dialed its public endpoint in Oct 2026, with no key and no account of yours. The names are the server’s own.
most_recent_not_vulnerableis_vulnerablelookup
Signals
- GitHub stars
- 1
- Last commit
- Feb 2026
Advanced
- Delivery
- fetter-mcp MCP server → your Ahel connector (mcp.ahel.ai) → your AI.
- Item type
- mcp-server
- Key
io-github-fetter-io-fetter-mcp- Source
- github.com/fetter-io/fetter-mcp
- Hosted endpoint
https://mcp.fetter.io/mcp
github.com/fetter-io/fetter-mcp
More in Security
MCP server
More in Securityechelongraph-mcp
MCP server · echelongraph
More in Securitynotfair
MCP server · nowork-studio
More in Securityfeedmyagent
MCP server · makash
More in Securityopenwork
MCP server · different-ai
More in Securityosv-advisory-mcp-server
MCP server · cyanheads
More in Security