# Vdb

> Vdb is a free MCP server for security. Check packages for CVEs, slopsquatting and CISA KEV before your agent installs them. It can run locally on your machine or as a hosted remote server.

- **Kind:** MCP server
- **Pricing:** Free
- **Platforms:** CLI, API
- **Website:** https://vdb.ai.kr
- **Repository:** https://github.com/0pstech/vdb-mcp
- **MCP Registry name:** kr.ai.vdb/vdb
- **Categories:** security, developer tools
- **Listing updated:** 2026-11-14

## About

Coding agents invent or pick package names faster than advisories appear, and this server gives them a verdict first. It flags known CVEs, names that an LLM hallucinated and an attacker may have registered (slopsquatting), CISA KEV status and trust tiers for community MCP servers, with names only and no source sent.

The hosted endpoint offers seven tools, including a lockfile or SBOM merge verdict that covers transitive dependencies. Running it locally with uvx adds four hardening tools that need your files: a call-site analysis of whether attacker data can reach a dangerous sink through dependencies, remediation ranking, post-fix verification with signed evidence and OpenVEX output; source never leaves the machine.

Package checks work without a key on a per-address allowance (100 packages an hour); other tools need a free key, currently unmetered beyond abuse protection. It must run where the code is, such as Claude Code, Cursor, Cline or Codex CLI, since a chat-only web AI has no outbound network or file access.

## Features

- **Slopsquatting detection:** Flags hallucinated package names that someone has registered.
- **CVE and KEV status:** Reports vulnerabilities and CISA known-exploited status.
- **Lockfile verdict:** Scans a whole lockfile or SBOM including transitives.
- **MCP server trust:** Gives a trust tier and permission scopes for community MCP servers.
- **Call-site hardening:** Local tools check whether attacker data reaches dangerous calls.
- **OpenVEX output:** Lists advisories that cannot be reached.

## Good for

- Gate every package an agent wants to install
- Scan a lockfile before merging
- Check an MCP server's scopes before adding it
- Rank which dependency fix removes the most risk

## Install

### Claude Code (Terminal)

```sh
claude mcp add --transport http vdb https://vdb.ai.kr/mcp
```

### Claude Desktop (Settings → Connectors → Add custom connector)

```sh
https://vdb.ai.kr/mcp
```

### Cursor (~/.cursor/mcp.json)

```json
{
  "mcpServers": {
    "vdb": {
      "url": "https://vdb.ai.kr/mcp"
    }
  }
}
```

### VS Code (.vscode/mcp.json)

```json
{
  "servers": {
    "vdb": {
      "type": "http",
      "url": "https://vdb.ai.kr/mcp"
    }
  }
}
```

### Windsurf (~/.codeium/windsurf/mcp_config.json)

```json
{
  "mcpServers": {
    "vdb": {
      "serverUrl": "https://vdb.ai.kr/mcp"
    }
  }
}
```

### Other clients (your client's MCP settings (most use this shape))

```json
{
  "mcpServers": {
    "vdb": {
      "type": "http",
      "url": "https://vdb.ai.kr/mcp"
    }
  }
}
```

## Tools

- `vdb_check_package`: Check one package for vulnerabilities, slop risk and KEV.
- `vdb_check_packages`: Bulk slopsquatting and risk check for a list of packages.
- `vdb_scan_lockfile`: Merge verdict for a whole lockfile or SBOM.
- `vdb_lookup`: Fetch one advisory by ID.
- `vdb_search`: Free-text search over the vulnerability corpus.
- `vdb_check_mcp_server`: Trust tier and permission scopes of a community MCP server.
- `vdb_list_slopsquatting`: Current slopsquatting candidates per ecosystem.
- `vdb_harden`: Call-site hardening analysis (local only).
- `vdb_harden_verify`: Re-analyze after a fix with signed evidence (local only).
- `vdb_harden_remediate`: Rank what to fix first (local only).
- `vdb_vex`: OpenVEX document of unreachable advisories (local only).

## FAQ

### Do I need a key?

Package checks work without one on a per-IP allowance; other tools need a free key.

### Does my source code leave my machine?

The package checks send names only, and the hardening tools run locally.

### Where does it work?

In an agent on your machine with the project folder, such as Claude Code, Cursor, Cline or Codex CLI.

### How do I install locally?

Run uvx vdb-mcp or pipx run vdb-mcp.

### Is Vdb free?

Yes. Vdb is free to use.

### How do I add Vdb to Claude Code?

Run this command in your terminal: claude mcp add --transport http vdb https://vdb.ai.kr/mcp

### What are the best alternatives to Vdb?

The closest alternatives to Vdb are REA, Codebase Memory, GitHub, Serena MCP: the IDE for your agent and XERJ.

## Alternatives

- [REA](https://ai-powered.com/mcp/rea): REA is an open-source MCP server for developer tools.
- [Codebase Memory](https://ai-powered.com/mcp/codebase-memory): Codebase Memory is an open-source MCP server for knowledge and memory.
- [GitHub](https://ai-powered.com/mcp/github-2): GitHub is an open-source MCP server for developer tools.
- [Serena MCP: the IDE for your agent](https://ai-powered.com/mcp/serena-mcp-the-ide-for-your-agent): Serena MCP: the IDE for your agent is an open-source MCP server for developer tools.
- [XERJ](https://ai-powered.com/mcp/xerj): XERJ is an open-source MCP server for knowledge and memory.
- [Ouroboros](https://ai-powered.com/mcp/ouroboros): Ouroboros is an open-source MCP server for developer tools.
- [Claude Code Ultimate Guide](https://ai-powered.com/mcp/claude-code-ultimate-guide): Claude Code Ultimate Guide is an open-source MCP server for ai models and llm tooling.
- [Concord MCP](https://ai-powered.com/mcp/concord-mcp): Concord MCP is a freemium MCP server for developer tools.

---
Source: https://ai-powered.com/mcp/vdb
