Integrate professional security analysis into Claude, GitHub Copilot, Cursor, and other AI agents. This MCP server allows your LLM to scan code, detect vulnerabilities, and use the SynapseCortex engine to fix them.

Features

Security Scanning: Scan code snippets, files, or entire directories via LLM commands.
AI Analysis: Explain vulnerabilities and generate secure fixes using the Cortex engine.
Cloud Sync: Upload findings to your SynapseAudit Dashboard.
Context Awareness: Provides LLMs with direct access to security patterns and CWE definitions.

Installation

1. Build from Source

# Clone repository
git clone https://github.com/digidenone/SynapseAudit.git
cd SynapseAudit/mcp-server

# Install and Build
npm install
npm run build

2. Configure Claude Desktop

Add the following to your claude_desktop_config.json:

{
  "mcpServers": {
    "synapseaudit": {
      "command": "node",
      "args": ["/absolute/path/to/SynapseAudit/mcp-server/dist/index.js"]
    }
  }
}

Available Tools

The server exposes the following tools to the LLM:

Tool	Description	Arguments
`scan_project`	Scan a directory for vulnerabilities.	`path` (string), `deep` (boolean)
`scan_sca`	Check dependencies for known vulnerabilities.	`path` (string)
`scan_secrets`	Find exposed API keys and credentials.	`path` (string)
`scan_iac`	Audit Infrastructure-as-Code files.	`path` (string)
`scan_sbom`	Generate CycloneDX SBOM.	`path` (string)
`analyze_code`	Analyze a specific code snippet.	`code` (string), `language` (string)
`explain_vuln`	Get a detailed explanation of a vulnerability type.	`vuln_id` (string)
`get_fix`	Generate a secure fix for a finding.	`vuln_id` (string), `context` (string)
`check_compliance`	Check code against specific standards (OWASP, GDPR).	`standard` (string)