Prompt Shield & AI Safety MCP

Deterministic prompt injection classification, instruction stripping, secret entropy detection, PII obfuscation, and dataset bias auditing.

Part of the WHAM ENGINE fleet of deterministic AI agent micro-services with automated Base L2 HTTP 402 (x402) USDC micropayments.

Tools Included in this Suite (5 Tools)

• prompt_injection_jailbreak_classifier: High-speed deterministic classifier detecting indirect prompt injections, delimiter hijacking, and system role impersonation attacks in user inputs.
• strip_prompt_injection: Neutralizes indirect prompt injections, system prompt leak probes, jailbreak tokens, and hidden instruction tags in untrusted text.
• secrets_entropy_scanner: Shannon entropy and pattern analyzer scanning source code for leaked private keys, AWS access secrets, JWTs, and database connection strings.
• obfuscate_pii_entities: Detects and masks Personally Identifiable Information (SSNs, credit cards, emails, phone numbers, API keys) prior to model ingestion.
• synthetic_dataset_bias_auditor: Audits synthetic agent training data distributions for demographic bias, representation skew, and label drift across sensitive attribute categories.

Recommended Workflow

An enterprise agent receives untrusted user uploads -> runs secrets entropy scan to catch leaked tokens -> obfuscates PII -> passes through the prompt injection classifier -> strips residual jailbreak tags before invoking downstream frontier models.

Quickstart Configuration

Add this server to your claude_desktop_config.json:

{
  "mcpServers": {
    "prompt-shield-security-mcp": {
      "command": "npx",
      "args": ["-y", "@whambammy/prompt-shield-security-mcp"],
      "env": {
        "PAYMENT_WALLET": "0x9793E7269b3301893318dEa8338576Ba612F39B3",
        "BASE_RPC_URL": "https://mainnet.base.org"
      }
    }
  }
}

Or connect the skill directly via Smithery:

smithery skill add whambammy/prompt-shield-security-mcp