Skip to content
MCP server

Valca

Read-only security scanner for code, infrastructure files and AI agent config

Runs locallyStatus and history
Pricing
Free
Platforms
CLI, vscode, Linux
Runs
Locally
Version
0.6.1
Listing updated
1 day ago

About Valca

Valca is a free MCP server for security. It runs locally on your machine and is started by your AI client.

Valca is a security scanner built to catch the insecure defaults AI coding assistants tend to reproduce. The headline example is a docker-compose port mapping such as 5432:5432, which binds to 0.0.0.0 and, because Docker writes its own firewall rules, bypasses UFW. The README says Checkov, Trivy, Snyk and Semgrep miss this, and Valca flags it with the fix of binding to 127.0.0.1.

It ships 116 rules in 27 categories. With Claude Code a hook scans every file before it is saved and blocks the write on critical or high findings, while a scan command, SARIF output and exit codes support CI. This MCP server lets agents call the scanner themselves. It is read-only and never returns the matched line, so a credential it finds does not enter a model's context.

The core engine is offline and stdlib-only Python. Three dependency rules send package names and versions to OSV, PyPI and npm, and can be disabled. Valca is under the Business Source License 1.1, free for production use except as a competing hosted service, and each release converts to MIT after four years.

Install

Run in a terminal

claude mcp add valca -- uvx valca

Add to claude_desktop_config.json

{
  "mcpServers": {
    "valca": {
      "command": "uvx",
      "args": [
        "valca"
      ]
    }
  }
}
Install in CursorOpens Cursor, which asks you to confirm. Or add it by hand:

Add to ~/.cursor/mcp.json

{
  "mcpServers": {
    "valca": {
      "command": "uvx",
      "args": [
        "valca"
      ]
    }
  }
}
Install in VS CodeOpens VS Code, which asks you to confirm. Or add it by hand:

Add to .vscode/mcp.json

{
  "servers": {
    "valca": {
      "type": "stdio",
      "command": "uvx",
      "args": [
        "valca"
      ]
    }
  }
}

Add to ~/.codeium/windsurf/mcp_config.json

{
  "mcpServers": {
    "valca": {
      "command": "uvx",
      "args": [
        "valca"
      ]
    }
  }
}

Add to your client's MCP settings (most use this shape)

{
  "mcpServers": {
    "valca": {
      "command": "uvx",
      "args": [
        "valca"
      ]
    }
  }
}

What it does

Generation-time blocking
Stops insecure writes before the file reaches disk.
116 rules
Covers 27 categories including IaC and agent configuration.
Read-only MCP
Agents can scan but never edit, and matched lines are withheld.
CI output
Exit codes and SARIF feed code-scanning gates.
Offline core
The engine sends no code anywhere; only optional dependency checks use the network.
Scan history
Local log and stats show rule frequency and suppression rates.

Good for

  • Block an exposed database port in docker-compose
  • Scan an AI-written project before committing
  • Feed SARIF findings to GitHub code scanning
  • Check dependencies for known CVEs or fabricated packages

Pricing

Free

Free for production use under BSL 1.1

Prices change. Check valca.io before you buy.

Valca FAQ

Does it send my code anywhere?

No. Only package names and versions go out for three dependency rules, which can be disabled.

What licence is it?

Business Source License 1.1, which the project says is not OSI open source.

Can the MCP server edit files?

No, it is read-only.

What happens on a critical finding?

The scan exits with code 2 and Claude Code blocks the write.

Is Valca free?

Yes. Valca is free to use.

How do I add Valca to Claude Code?

Run this command in your terminal: claude mcp add valca -- uvx valca

What are the best alternatives to Valca?

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

Alternatives to Valca

  1. REA

    MCP server

    CLI and MCP server that lets coding agents reverse engineer apps and binaries

    Runs locally
    • Open source
    • 464K GitHub stars
    • Developer tools
  2. Codebase Memory

    MCP server

    Local MCP server that builds a code knowledge graph for AI coding agents

    Runs locally
    • Open source
    • 48.7K GitHub stars
    • Knowledge and memory
  3. GitHub

    MCP server

    Official MCP server connecting AI assistants to GitHub repos, issues and workflows

    Not checked yet
    • Remote
    • Open source
    • 34.5K GitHub stars
    • Developer tools
  4. MCP toolkit giving coding agents symbol-level code retrieval and editing

    Runs locally
    • Open source
    • 31.5K GitHub stars
    • Developer tools
  5. XERJ

    MCP server

    Local search engine for AI agents with cited answers over your code and docs

    Runs locally
    • Open source
    • 19.2K GitHub stars
    • Knowledge and memory
  6. Ouroboros

    MCP server

    Spec-first layer that pins requirements and verifies results of AI coding agent work

    Runs locally
    • Open source
    • 6.4K GitHub stars
    • Developer tools