MCP Scorecard
Deterministic CI scanner and surface-risk scoring for MCP (Model Context Protocol) servers.
What is MCP Scorecard?
MCP Scorecard is a Model Context Protocol (MCP) server that allows AI assistants like Claude, Cursor, and VS Code to deterministic ci scanner and surface-risk scoring for mcp (model context protocol) servers.
Deterministic CI scanner and surface-risk scoring for MCP (Model Context Protocol) servers.
This server falls under the Security category on MCPgee, the world's largest MCP server directory with 33,000+ servers.
Features
- Deterministic CI scanner and surface-risk scoring for MCP (M
Use Cases
Maintainer
Works with
Installation
Manual Installation
npx mcp-scorecardConfiguration
Configuration Details
claude_desktop_config.json
Performance
Response Metrics
Resource Usage
How to Set Up and Use MCP Scorecard
MCP Scorecard is a deterministic CI scanner and risk-scoring tool for Model Context Protocol servers, designed to evaluate MCP server implementations across four dimensions: conformance, security, ergonomics, and metadata completeness. It generates a composite trust score (0-100) and can produce JSON and SARIF output for integration into CI/CD pipelines such as GitHub Actions. Security teams and MCP server authors use it to identify dangerous tool patterns, missing schema definitions, and unclear destructive capability declarations before deploying MCP servers to production.
Prerequisites
- Python 3.8 or later installed
- pip package manager
- The MCP server you want to scan must be runnable as a local command
- Optional: GitHub Actions for automated CI scanning
Clone the repository
Clone the MCP Scorecard repository and navigate into it.
git clone https://github.com/aak204/MCP-Scorecard.git
cd MCP-ScorecardSet up a virtual environment
Create and activate a Python virtual environment.
python -m venv .venv
source .venv/bin/activate # On Windows: .venv\Scripts\activateInstall MCP Scorecard
Install the package and its development dependencies.
pip install -e .[dev]Run a basic scan
Scan an MCP server by providing the command used to start it. The scanner will connect to the server, enumerate its tools, and compute a score.
mcp-scorecard scan --cmd python examples/insecure-server/server.pyRun with scoring gate and output artifacts
Use --min-score to fail CI if the score falls below a threshold, and --json-out / --sarif to save structured reports for downstream processing or GitHub Security tab integration.
mcp-scorecard scan \
--min-score 80 \
--json-out mcp-scorecard-report.json \
--sarif mcp-scorecard-report.sarif \
--cmd python your_server/server.pyIntegrate into GitHub Actions
Add MCP Scorecard as a step in your CI workflow using the provided GitHub Action.
- name: MCP Scorecard
uses: aak204/[email protected]
with:
cmd: python your_server/server.py
min-score: 80MCP Scorecard Examples
Client configuration
MCP Scorecard is a CLI scanning tool, not a running MCP server. Below is a GitHub Actions workflow snippet showing how to run it in CI.
{
"name": "MCP Security Scan",
"on": ["push", "pull_request"],
"jobs": {
"scorecard": {
"runs-on": "ubuntu-latest",
"steps": [
{"uses": "actions/checkout@v4"},
{"name": "Run MCP Scorecard",
"uses": "aak204/[email protected]",
"with": {
"cmd": "python server.py",
"min-score": "80"
}
}
]
}
}
}Prompts to try
CLI commands to use when auditing MCP servers with MCP Scorecard.
- "mcp-scorecard scan --cmd python my_server/server.py" — Basic scan with console output
- "mcp-scorecard scan --min-score 75 --json-out report.json --cmd npx my-mcp-server" — Scan with JSON report and score gate
- "mcp-scorecard scan --sarif report.sarif --cmd python server.py" — Generate SARIF for GitHub Security tab
- "mcp-scorecard scan --cmd docker run --rm my-mcp-image" — Scan a containerized serverTroubleshooting MCP Scorecard
Server command fails to start during scan
Ensure the MCP server command is fully executable in isolation before scanning. Test it manually with the same command string passed to --cmd. Check that all dependencies are installed in the active environment.
Score is unexpectedly low due to security flags
Review the JSON output's security section to see which patterns triggered risk flags. Common causes are tools that execute shell commands, write to arbitrary filesystem paths, or make outbound network requests without declaring this in their descriptions.
SARIF output not appearing in GitHub Security tab
After generating the SARIF file, you must upload it using the actions/upload-sarif action with category: mcp-scorecard. Ensure your repository has GitHub Advanced Security enabled for the Security tab to display SARIF results.
Frequently Asked Questions about MCP Scorecard
What is MCP Scorecard?
MCP Scorecard is a Model Context Protocol (MCP) server that deterministic ci scanner and surface-risk scoring for mcp (model context protocol) servers. It connects AI assistants to external tools and data sources through a standardized interface.
How do I install MCP Scorecard?
Follow the installation instructions on the MCP Scorecard GitHub repository. Clone the repo, install dependencies, and add the server config to your AI client.
Which AI clients work with MCP Scorecard?
MCP Scorecard works with all major MCP-compatible AI clients including Claude Desktop, Claude Code, Cursor, VS Code (GitHub Copilot), Windsurf, and Cline.
Is MCP Scorecard free to use?
Yes, MCP Scorecard is open source and available under the Apache-2.0 license. You can use it freely in both personal and commercial projects.
MCP Scorecard Alternatives — Similar Security Servers
Looking for alternatives to MCP Scorecard? Here are other popular security servers you can use with Claude, Cursor, and VS Code.
Casdoor
★ 13.6kAn open-source Agent-first Identity and Access Management (IAM) /LLM MCP & agent gateway and auth server with web UI supporting OpenClaw, MCP, OAuth, OIDC, SAML, CAS, LDAP, SCIM, WebAuthn, TOTP, MFA, Face ID, Google Workspace, Azure AD
ghidraMCP
★ 9.0kAn Model Context Protocol server that enables LLMs to autonomously reverse engineer applications by exposing Ghidra's decompilation and analysis tools. It allows AI agents to list code structures, rename methods, and analyze binaries directly through
HexStrike AI
★ 8.9kHexStrike AI MCP Agents is an advanced MCP server that lets AI agents (Claude, GPT, Copilot, etc.) autonomously run 150+ cybersecurity tools for automated pentesting, vulnerability discovery, bug bounty automation, and security research. Seamlessly b
IDA Pro MCP
★ 8.7kEnables AI-assisted reverse engineering in IDA Pro by providing tools to analyze binaries, decompile functions, manage comments, search patterns, and interact with the IDA database through natural language.
Anthropic Cybersecurity Skills
★ 6.6k754 structured cybersecurity skills for AI agents · Mapped to 5 frameworks: MITRE ATT&CK, NIST CSF 2.0, MITRE ATLAS, D3FEND & NIST AI RMF · agentskills.io standard · Works with Claude Code, GitHub Copilot, Codex CLI, Cursor, Gemini CLI & 20+ platform
Hooker
★ 5.1k🔥🔥 hooker is a Frida-based reverse engineering toolkit for Android. It offers a user-friendly CLI, universal scripts, auto hook generation, memory roaming to detect activities/services, one-click SOCKS5 proxy setup, Frida JustTrustMe, and BoringSSL u
Browse More Security MCP Servers
Explore all security servers available in the MCPgee directory. Each server includes setup guides for Claude, Cursor, and VS Code.
Set Up MCP Scorecard in Your Editor
Choose your AI client for step-by-step setup instructions.
Quick Config Preview
Add this to your claude_desktop_config.json or .cursor/mcp.json
Ready to use MCP Scorecard?
Browse our complete directory of 33,000+ MCP servers, read setup guides for your editor, and start building with the Model Context Protocol.