VulX Watch

ai.vulxPublished by vulx.aiOfficial publisherv1.0.0Updated Oct 24, 2026

Independent security review for AI-built apps. Watch a GitHub repo. Never a patch.

VerifiedStreamable HTTPWeb executableDeveloper ToolsSecurity & Monitoring

Overview

AI-generated overview

VulX Watch is a remote MCP connector that lets an assistant register a GitHub repo for independent security review and report what the latest reading found.

What you can do
  • Watch a GitHub repository through the VulX Watch connector
  • Ask what the most recent security reading found for a watched repo
  • Check security status through the watch-repo and security-status commands
  • Read committed source only; it does not patch code, open PRs, or log in to a running site or database
When to use it
Useful when you build apps with AI assistance and want an independent security review of a GitHub repository's committed source, and want the agent to be able to report the latest findings on request.
Requirements
A remote MCP endpoint at mcp.vulx.ai over streamable HTTP, plus a VulX account; OAuth sign-in is required and happens when the client connects. No environment variables or headers are declared. The README notes it is free while in development.
Before you install
Requires OAuth sign-in to a VulX account and sends repository review data to VulX. It reads committed source, so private code is shared with the service. It writes its own brief files (AGENTS.md and CLAUDE.md) when brief delivery is on. It gives no score or ship/no-ship verdict and does not patch code, so findings need human judgment.

Installation

In SourceWeft

  1. Open VulX Watch in the dashboard and add it to a workspace.
  2. Connect the required credentials. They are stored privately and never shown on public pages.
  3. Enable the server for the chats that should use its tools.

Web executable via Streamable HTTP. Remote servers run from the web runtime once configured in a workspace.

Use with an agent

Paste this prompt into Claude Code, Codex, Cursor or another agent — or into SourceWeft chat. The agent reads this server's install guide, shows you its publisher, endpoint and sign-in, and adds it to your client once you agree. Keys never go into a command.

Read https://sourceweft.com/mcp/ai.vulx%2Fwatch/install.md and connect the MCP server it describes to my client. Before changing any config, show me its publisher, endpoint and sign-in method and wait for my OK. Never put an API key in a command or in a file I would commit; ask me where it should come from.

Read the install guide the agent follows

Other MCP clients

Add this to your client's mcpServers config. The client completes OAuth sign-in on first connect.

{
  "mcpServers": {
    "vulx-watch": {
      "type": "http",
      "url": "https://mcp.vulx.ai/mcp"
    }
  }
}

README

VulX Watch for Cursor

Adds the VulX Watch connector to Cursor: independent security review for AI-built apps. The agent can watch a GitHub repo and ask what the last reading found.

VulX does not write a patch or open a PR.

What you get

  • MCP server at https://mcp.vulx.ai/mcp
  • A skill so the agent knows when to call Watch
  • Commands: watch-repo, security-status

You still need a VulX account. Sign-in happens when Cursor talks to the connector. Free while we build it.

Install

  1. Install VulX Watch from the Cursor Marketplace, or clone this repo into ~/.cursor/plugins/local and restart Cursor.
  2. Complete the VulX sign-in prompt.
  3. Ask: Watch my repo owner/name with VulX. Then: What did VulX find?

If OAuth does not finish inside Cursor, add the same URL as a custom connector: https://mcp.vulx.ai/mcp. Human page: https://vulx.ai/connect

What it will not do

  • Patch your app or open a merge-ready PR
  • Log in to your running site or database
  • Give a 0–100 score or a “safe to ship”
  • Treat a quiet week as an all-clear

It reads committed source. It writes only its own brief (AGENTS.md and CLAUDE.md) when brief delivery is on.

License

MIT

Source: README.md at commit becdf56

Tools

0
This server asks you to sign in, so SourceWeft could not read its tools without an account and has not indexed them. What a connected account sees can depend on that account and its permissions.

Version history

1
  1. v1.0.0LatestSep 16, 2026

Score

85% of the applicable checks met

How complete this listing is, computed from the registry entry, the endpoint and public sources. Not a rating of the server and not a statement that it is safe or working today: reachability is a dated check, and no installs or reviews go into the score.

  • Reached in the last 30 daysRequired25 pts
  • Auth declaredRequired10 pts
  • Tools listed15 pts

    Answer tools/list without a credential, or ship a package that starts in a sandbox without one.

  • DescriptionRequired15 pts
  • Logo10 pts
  • Documentation15 pts
  • License10 pts