
Securityscorecard Mcp
io.github.CallMarcusPublished by GitHub account CallMarcusv2.0.0Updated Oct 26, 2026
Community-built, comprehensive MCP server for the SecurityScorecard API (unofficial).
Overview
Unofficial MCP server that lets an assistant query SecurityScorecard security ratings, risks, assets and email security for a company domain.
- What you can do
- Read security posture: scores, grades, key metrics and issue-type breakdowns
- Analyze risks and build remediation improvement plans
- Inventory assets and check SPF/DMARC/DKIM email security
- Search the SecurityScorecard API reference and run direct API queries
- Verify data completeness across tools, with minimal, standard or detailed response modes
- When to use it
- Useful when you want an assistant to pull third-party security ratings and risk context for a company domain, review email authentication posture, or draft remediation priorities without leaving the chat client.
- Requirements
- Runs locally over stdio via npx, so Node.js 20+ is needed. Requires a SecurityScorecard API token in SECURITY_SCORECARD_API_TOKEN; COMPANY_DOMAIN is an optional default domain. Network access to the SecurityScorecard API is required.
Installation
In SourceWeft
- Open Securityscorecard Mcp in the dashboard and add it to a workspace.
- Enable the server for the chats that should use its tools.
Desktop only via STDIO. STDIO servers start a local process, so they need the SourceWeft desktop host.
Other MCP clients
Follow the launch instructions in the repository.
README
SSC MCP Server
A community-built, comprehensive Model Context Protocol (MCP) server that integrates with the SecurityScorecard API. It runs over stdio, so it works with any MCP-compatible client — Claude Desktop, Claude Code, Cursor, VS Code, and others. It serves MCP protocol revision 2026-07-28 and stays compatible with 2025-era clients.
Published on npm as
@callmarcus/securityscorecard-mcpand listed in the MCP Registry asio.github.CallMarcus/securityscorecard-mcp.
Disclaimer: This is an independent, community-built open-source project. It is not affiliated with, endorsed by, sponsored by, or associated with SecurityScorecard, Inc. in any way. It is built solely against SecurityScorecard's publicly available API documentation. "SecurityScorecard" and all related names, marks, and logos are trademarks of SecurityScorecard, Inc. and are used here for identification purposes only. You must supply your own API credentials and comply with SecurityScorecard's terms of service.
Quick Start
Prerequisites
- Node.js 20+ - Download
- SecurityScorecard API Token - Get from your SecurityScorecard dashboard
Option A — Install from npm (recommended)
No clone or build required. The server runs over stdio via npx, so any MCP-compatible client can launch it. npx -y always fetches the latest published version.
Most clients — Claude Desktop, Cursor, Cline, Windsurf, and others — share the same mcpServers JSON. Add this block to the client's MCP config:
Where that config file lives:
Replace the credentials with your own, then restart the client.
Claude Code — add it from the CLI instead:
On Windows, wrap the launcher in cmd /c: ... -- cmd /c npx -y @callmarcus/securityscorecard-mcp.
VS Code (Copilot) — uses a servers key with an explicit type, in .vscode/mcp.json:
Option B — Run from source (for development)
Then point your MCP client at the local build. For clients that use the mcpServers format (Claude Desktop, Cursor, …):
Important: Replace the path and credentials with your actual values, then restart your MCP client. (For Claude Code, run claude mcp add security-scorecard --env SECURITY_SCORECARD_API_TOKEN=your-api-token-here -- node /path/to/security-scorecard-mcp/build/index.js.)
Available Tools
The server (index.js) provides 9 specialized tools:
Response Modes
Each tool supports three response modes for token efficiency:
- minimal - Quick answers (15-50 tokens)
- standard - Overview with context (200-300 tokens)
- detailed - Comprehensive analysis (800+ tokens)
Environment Variables
Optional rate limiting and caching:
API Discovery
The server includes hybrid search (semantic + keyword) for finding SecurityScorecard API endpoints:
This searches 535 indexed endpoints and returns matching paths with confidence scores, required parameters, and curl examples.
To update the API reference after changes:
Development
Build Commands
Project Structure
Testing
Troubleshooting
Build fails with out of memory
Use the fast build instead:
"Cannot find module" errors
Reinstall dependencies:
Semantic search degrades to keyword-only (Windows + WSL)
Install for the platform that runs the server. Claude Desktop on Windows
launches the server with Windows node, so if npm install ran under WSL
the native modules (onnxruntime-node, sharp) only have linux binaries —
the embeddings layer fails to load and api_discovery silently degrades to
keyword-only search (results still come back, but confidence scoring is
cruder). Run npm install && npm run build:fast from PowerShell or cmd in
the repo directory instead — or keep two clones, one per platform.
Your client doesn't see the server
- Double-check the config file location for your client (see Quick Start)
- For a from-source install, verify the path to
build/index.jsis correct - Restart the client completely
- Sanity-check that the server starts on its own:
npx -y @callmarcus/securityscorecard-mcp(it should launch and wait silently on stdio)
API returns 401 Unauthorized
Your API token is invalid or expired. Get a new one from SecurityScorecard dashboard.
License
MIT
Links
Source: README.md at commit af4fad8
Tools
0Version history
2- v2.0.0LatestOct 4, 2026
- v1.2.0Sep 16, 2026
Score
40% of the applicable checks met
How complete this listing is, computed from the registry entry, the endpoint and public sources. Not a rating of the server and not a statement that it is safe or working today: reachability is a dated check, and no installs or reviews go into the score.
- Reached in the last 30 days25 pts
Keep the endpoint answering the MCP handshake; SourceWeft probes it every 30 days.
- Auth declared10 pts
Declare the credential headers in the registry entry, or keep the endpoint answering the probe so its auth can be classified.
- Tools listed15 pts
Answer tools/list without a credential, or ship a package that starts in a sandbox without one.
- DescriptionRequired15 pts
- Logo10 pts
- Documentation15 pts
- License10 pts
Declare a license in the repository or the registry entry.