Skip to content
MCP server

Agent-Shield

Scan text, tool output and web pages for prompt injection, secrets, PII and SSRF URLs

Install
Pricing
Freemium
Platforms
API, CLI, self-hosted
Runs
Remotely
Version
0.2.1
Listing updated
3 days ago

About Agent-Shield

Agent-Shield is a freemium MCP server for security. It runs as a hosted remote server, so there is nothing to install locally.

Agent-Shield sits before an agent acts. The check is deterministic, so the same input gives the same result, and content is processed in memory and not stored.

The maker is unusually frank about limits. Its SSRF detection reports 77% recall on its dev set and 62% on a blind holdout of 200 samples. The README says recall on novel prompt injection is only 4.7% on an external hold-out, and no_signals is not a safety guarantee. Treat it as a first layer.

A key comes from one API call with no email or card, and the free tier allows 1,000 scans a day. Paid tiers listed are Pro $20, Power $200 and Team $500 a month. The hosted MCP endpoint offers shield.scan, and a Python stdio wrapper exposes shield_scan.

Install

Run in a terminal

claude mcp add --transport http agent-shield https://agent-shield.startekenterprises.com/mcp

Add to Settings → Connectors → Add custom connector

https://agent-shield.startekenterprises.com/mcp
Install in CursorOpens Cursor, which asks you to confirm. Or add it by hand:

Add to ~/.cursor/mcp.json

{
  "mcpServers": {
    "agent-shield": {
      "url": "https://agent-shield.startekenterprises.com/mcp"
    }
  }
}
Install in VS CodeOpens VS Code, which asks you to confirm. Or add it by hand:

Add to .vscode/mcp.json

{
  "servers": {
    "agent-shield": {
      "type": "http",
      "url": "https://agent-shield.startekenterprises.com/mcp"
    }
  }
}

Add to ~/.codeium/windsurf/mcp_config.json

{
  "mcpServers": {
    "agent-shield": {
      "serverUrl": "https://agent-shield.startekenterprises.com/mcp"
    }
  }
}

Add to your client's MCP settings (most use this shape)

{
  "mcpServers": {
    "agent-shield": {
      "type": "http",
      "url": "https://agent-shield.startekenterprises.com/mcp"
    }
  }
}

Tools this server provides (1)

Taken from the project documentation; the running server may differ.

shield.scan
Scan content for secrets, PII, risky URLs and injection patterns

What it does

Prompt injection screening
Flags known injection patterns.
Secrets and PII
Finds leaked credentials and personal data.
SSRF risk URLs
Flags cloud metadata, loopback and private targets.
Signed receipts
Each scan returns a receipt.
Published benchmarks
Recall figures, misses included.

Good for

  • Scan fetched web content before an agent follows it
  • Check tool output for leaked secrets
  • Screen emails for hidden instructions
  • Flag unsafe URLs before a request

Pricing

Freemium, from $20 a month

Free 1,000 scans/day; Pro $20, Power $200, Team $500 per month

Prices change. Check agent-shield.startekenterprises.com before you buy.

Agent-Shield FAQ

How do I get a key?

One POST to /v1/signup, no email or card.

What is the free limit?

1,000 scans a day.

Is it a guarantee?

No. It is described as a first layer, weak on novel prompt injection.

Is data stored?

The site says content is processed in memory and never stored.

Is Agent-Shield free?

Agent-Shield has a free plan, with paid plans for more usage or features. Paid plans start at $20 a month.

How do I add Agent-Shield to Claude Code?

Run this command in your terminal: claude mcp add --transport http agent-shield https://agent-shield.startekenterprises.com/mcp

What are the best alternatives to Agent-Shield?

The closest alternatives to Agent-Shield are REA, Agent Roast, Midplane, Leash and Tinplate Edge AI Safety, Security & EU Compliance.

Alternatives to Agent-Shield

  1. REA

    MCP server

    CLI and MCP server that lets coding agents reverse engineer apps and binaries

    Runs locally
    • Open source
    • 464K GitHub stars
    • Developer tools
  2. Agent Roast

    MCP server

    Offline review that flags agent changes which weaken tests, assertions or CI

    Runs locally
    • Open source
    • 4 Hacker News points
    • Developer tools
  3. Midplane

    MCP server

    Postgres gateway for AI agents with table grants, column masking and approval for writes

    Runs locally
    • Open source
    • 5 GitHub stars
    • Databases
  4. Leash

    MCP server

    Review what your AI touched, approve risky actions and undo supported changes

    • 1 tool
    • Remote
    • Free
    • 1 GitHub star
    • Security
  5. SecureCloud

    MCP server

    SentinelOne and IT security catalog, prices, quotes and checkout links for Israel

    • 7 tools
    • Remote
    • Free
    • Security