ShieldAPI MCP (shieldapi-mcp)
@alberthild
MCP server providing 9 security intelligence tools for AI agents. Real-time threat detection with pay-per-request USDC micropayments via x402 protocol. Free tier included (3 calls/day, no account needed).
Overview
What is Shieldapi Mcp?
Shieldapi Mcp provides 9 security tools via the Model Context Protocol, including password breach checks, prompt injection detection, and domain/IP reputation lookups. It is designed for AI agents and developers who need real‑time security intelligence, and all tools work immediately in a free demo mode without any API key or wallet.
How to use Shieldapi Mcp?
Run npx shieldapi-mcp to start the MCP server. All tools are available in demo mode out of the box; to use paid mode, add a wallet for x402 USDC micropayments on Base.
Key features of Shieldapi Mcp
- Check passwords against 900M+ breached hashes (HIBP)
- Prompt injection detection with 200+ patterns under 100ms
- AI skill/plugin supply chain scanner with 8 risk categories
- Pay-per-request USDC micropayments via x402 on Base
- No API key required; works instantly in demo mode
- Includes email, domain, IP, and URL reputation checks
Use cases of Shieldapi Mcp
- Securing AI agents against prompt injection and jailbreak attacks
- Auditing AI plugins and skills for malicious code or credential leaks
- Verifying user passwords and emails against real breach databases
- Checking domain, IP, and URL reputation for phishing or malware
- Performing a combined full security scan in a single API call
FAQ from Shieldapi Mcp
What is the pricing model?
Each tool has a fixed per‑request price in USDC (e.g., check_password $0.001, check_prompt $0.005). A free demo mode is available with no payment required.
Do I need an API key or wallet to start?
No. The server works out of the box in demo mode without any API key or wallet. For paid mode you add a wallet for x402 USDC micropayments on Base.
What data sources does Shieldapi Mcp use?
It uses Have I Been Pwned (HIBP) for password and email breaches, DNS/SPF/DMARC/SSL data for domains, and blacklists, Tor exit detection, and reverse DNS for IPs. The skill scanner is based on the Snyk ToxicSkills taxonomy.
How comprehensive is the prompt injection detection?
It covers 200+ patterns including Base64, ROT13, Unicode homoglyphs, DAN/jailbreak prompts, and exfiltration attempts.
Is there a combined scan tool?
Yes, the full_scan tool runs all checks in one call for $0.010 per request.
