VulnFeed

Dependency vulnerability scanner with EPSS scoring. 9 MCP tools. Free tier + x402.

OtherPythonv0.3.3

VulnFeed — Dependency Vulnerability Monitoring for Claude Code

CI PyPI License: MIT vulnfeed-mcp MCP server

An MCP server that scans your project dependencies for known vulnerabilities, enriches with EPSS exploit probability scores, and recommends fix versions.

Free tier — 10 scans/day, 1 monitored project, no signup required. Single-CVE lookup and webhooks need a license key or x402.

Homepage: vulnfeed.novadyne.ai

Install

uvx vulnfeed-mcp

MCP client config

Add to your MCP client config (~/.claude/settings.json for Claude Code, claude_desktop_config.json for Claude Desktop):

Free tier (no signup, no API key):

{
  "mcpServers": {
    "vulnfeed": {
      "command": "uvx",
      "args": ["vulnfeed-mcp"]
    }
  }
}

Paid ($14/mo, unlimited scans + projects):

{
  "mcpServers": {
    "vulnfeed": {
      "command": "uvx",
      "args": ["vulnfeed-mcp"],
      "env": {
        "VULNFEED_API_KEY": "YOUR_LICENSE_KEY_HERE"
      }
    }
  }
}

Get a license key at vulnfeed.novadyne.ai.

x402 micropayments

VulnFeed also accepts x402 micropayments — AI agents can pay per scan with USDC on Base, no API key or signup needed. When the free tier limit is reached, the API returns HTTP 402 with payment requirements that x402-compatible clients handle automatically.

  • $0.01 per scan
  • $0.002 per CVE lookup
  • $0.50 per project per 30 days of hourly webhook monitoring (renewable; the webhook receives an expiring event with the renew URL a day before)

Tools

Scanning

ToolDescription
scan_projectAuto-detect and scan all lockfiles in a directory
scan_lockfileScan a specific lockfile
check_packageCheck a single package for vulnerabilities
lookup_cveDetailed CVE info with EPSS + fix versions (license key or x402)

Monitoring

ToolDescription
monitor_projectRegister for continuous monitoring (optional webhook_url for push alerts)
check_alertsNew vulns since last scan
set_webhookSet, rotate or remove a project's webhook
update_depsUpdate snapshot after upgrading packages
list_monitoredSee all monitored projects
unmonitor_projectRemove from monitoring

Supported lockfiles

  • package-lock.json (npm)
  • yarn.lock (Yarn)
  • pnpm-lock.yaml (pnpm)
  • requirements.txt (pip)
  • Pipfile.lock (Pipenv)
  • go.sum / go.mod (Go)
  • Cargo.lock (Rust)
  • Gemfile.lock (Ruby)
  • composer.lock (PHP)

How it works

  1. Parses your lockfile to extract dependency names + versions
  2. Queries OSV.dev (GitHub Advisories plus each ecosystem's security database) for known CVEs
  3. Enriches with EPSS exploit probability scores
  4. Filters noise — suppresses low-EPSS, non-critical CVEs by default
  5. Sorts by exploitability — most likely to be exploited first
  6. Returns fix versions from each advisory's affected-version ranges

Smart filtering

By default, VulnFeed suppresses low-priority CVEs (EPSS < 10% AND CVSS < 9.0). This cuts noise by ~80%.

Pass show_all=True to any scan tool to see everything.

Continuous monitoring

  1. monitor_project — takes a baseline snapshot of current deps + known vulns
  2. check_alerts — diffs against baseline, surfaces only new vulns
  3. Either run check_alerts periodically, or pass webhook_url and let VulnFeed do it

With a webhook (a paid feature: a license key, or for agents x402 at $0.50 per project per 30-day term), VulnFeed re-scans the project every hour and POSTs new findings to your URL:

{
  "event": "vulnfeed.new_vulns",
  "delivery_id": "…", "project_id": "…", "project_name": "…", "sent_at": "…",
  "count": 1,
  "new_vulns": [{ "id": "GHSA-…", "package": "express", "version": "4.18.2", "severity": "HIGH",
                  "epss": { "score": 0.42 }, "fix_version": "4.19.2", "summary": "…" }],
  "alerts_url": "https://vulnfeed-api.novadyne.ai/vulnscan/alerts?project=…"
}

Every delivery carries X-VulnFeed-Signature: sha256=<hex>, the HMAC-SHA256 of the raw body under the webhook_secret returned once when the webhook was set. Verify it. Failed deliveries are retried on later sweeps for about a day, then dropped and counted. Rotate or remove the webhook with set_webhook.

License

MIT

Installation

Source-derived launch command. Check the maintainer’s required arguments and credentials before running:

bash
uvx vulnfeed-mcp

Set up in your AI client

Merge this template into ~/Library/Application Support/Claude/claude_desktop_config.json. Keep existing servers. Add any arguments, credentials, and permissions required by the maintainer; this template has not been install-tested.

json
{
  "mcpServers": {
    "io-github-infai-tech-vulnfeed": {
      "command": "uvx",
      "args": [
        "vulnfeed-mcp"
      ]
    }
  }
}

Restart Claude Desktop completely for changes to take effect. Confirm the server appears connected in the client’s tool list, then try a read-only example from its documentation.

Claude Desktop setup reference

Package

vulnfeed-mcppypi

Compatible MCP Clients

VulnFeed works with any MCP-compatible client. Copy the config snippet from the Configuration section above and add it to the file shown for your client, then restart the application.

  • Claude Desktop~/Library/Application Support/Claude/claude_desktop_config.jsonRestart Claude Desktop completely for changes to take effect.
  • Cursor~/.cursor/mcp.jsonRestart Cursor for changes to take effect.
  • VS Code.vscode/mcp.jsonReload VS Code window for changes to take effect.
  • Windsurf~/.codeium/windsurf/mcp_config.jsonRestart Windsurf for changes to take effect.
  • Claude Code.mcp.jsonSave at the project root, then start Claude Code in that project and review the MCP server approval prompt. Keep real credentials out of shared files.

Learn More