Works with the AI Agents you already use
MCP Config Auth Auditor
Audit MCP server configurations for authorization leaks, excessive tool scopes, and command injection risks.
$15
MCP Config Auth Auditor
Example session with this skill installed
Audit my claude_desktop_config.json for security risks. I have a postgres-mcp and a filesystem-mcp server running.
- Read your context and instructions
- Compiled the mcp config auth
I have completed the security audit of your claude_desktop_config.json. The analysis identified two high-risk configurations related to broad filesystem access and hardcoded database credentials.
MCP Configuration Security Audit
| Component | Risk Level | Finding | Recommendation |
|---|---|---|---|
filesystem-mcp | High | Root directory (/) access enabled. | Restrict to specific project subdirectories. |
postgres-mcp | Critical | Hardcoded password in connection string. | Use env variables or .env file references. |
shell-tool | Medium | Unquoted command arguments. | Wrap arguments in quotes to prevent injection. |
Hardened Configuration Template
{
"mcpServers": {
"postgres": {
"command": "npx",
"args": ["-y", "@modelcontextprotocol/server-postgres"],
"env": {
"DATABASE_URL": "env:MCP_POSTGRES_URL"
}
},
"filesystem": {
"command": "npx",
"args": ["-y", "@modelcontextprotocol/server-filesystem", "./safe_zone"]
}
}
}
Next steps
- Move your database connection string to a local
.envfile. - Create the
./safe_zonedirectory and move necessary files there. - Restart your MCP client to apply the scoped-down permissions.
Connects securely to your tools. The creator never sees your data.
About this skill
MCP configuration files often hide critical security flaws, from exposed API keys to overly permissive shell access. This skill audits your Model Context Protocol server setups to identify authorization leaks and command injection risks before they are exploited.
What it does
- Audit Tool Scopes: Evaluates filesystem, shell, and network tool registrations against the principle of least privilege.
- Detect Command Injections: Flags insecure execution vectors like unquoted paths and unrestricted bash tools.
- Identify Auth Leaks: Locates hardcoded tokens and insecure credential handling in configuration files.
- Generate Hardened Configs: Produces secure JSON templates with environment variable isolation.
How it works
- Analyze Configs: Provide a
claude_desktop_config.jsonor Cursor settings file for inspection. - Review Risks: The skill flags specific lines where permissions exceed operational necessity.
- Remediate: Review the generated hardened JSON output that replaces raw secrets with placeholders.
Frameworks & tools
Specifically designed for Model Context Protocol (MCP) implementations, Claude Desktop configurations, and Cursor IDE settings.
Why this beats prompting it yourself
Generic security prompts often miss MCP-specific nuances like tool-call delegation risks or the way different IDEs handle environment variables. This skill uses a deterministic workflow focused specifically on the MCP specification and least-privilege architecture.
Use cases
- Reviewing a new MCP server configuration for excessive filesystem permissions.
- Hardening production JSON configs to move secrets into environment variables.
- Checking third-party MCP tool registrations for command injection vulnerabilities.
Known limitations
Does not automatically modify system-level files or global paths. All outputs require manual implementation to prevent accidental system lockouts.
Find more skills like this on the Claude Skills Marketplace, the Codex Skills Marketplace and the Gemini Skills Marketplace, or browse Claude Skills for Security.
Restricted to project workspace; cannot access parent directories or system-level config paths. Does not automatically apply changes to live systems.
How to use
- 1
Connect your agent to the Agensi MCP
Copy the prompt and paste it into Claude, ChatGPT, Cursor or any agent that supports MCP. It sets up the connection for you. Manual setup
- 2
Sign in to Agensi
Your agent shows a sign-in link. Sign in and click Allow.
- 3
Ask for what you need
Your agent loads this skill and uses it on your task. It always asks before buying a skill.
Reviews
No reviews yet
Be one of the first to try it. Every listed skill goes through the trust checks below.
Security scanned
Checked by our 8-point scan before listing
Fresh listing
Recently published to Agensi
Money-back guarantee
Doesn't do what its page promises? You get your money back.
Trust & safety
Security scan
Security scan passed 4 days ago
- Software
- Config Auth Auditor MCP
File System Read/Write (strictly scoped to current project workspace `./`). Zero global installation or elevated privileges required.
Permissions
Target Use Cases: - Audit my MCP server configuration for security vulnerabilities. - Check Claude Desktop MCP settings for exposed credentials and shell risks. - Harden Model Context Protocol tool permissions.