Skip to main content
Glama
Prompt Shield & AI Safety MCP

Prompt Shield & AI Safety MCP

Prompt Shield & AI Safety MCP

Smithery Compatible Model Context Protocol Base L2 Settlement Tools License: MIT

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

Built specifically for LLM application builders, LangChain/LlamaIndex engineers, enterprise AI developers, and AI red-teamers.


⚑ Quickstart

Smithery Install

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

Claude Desktop / Cursor (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"
      }
    }
  }
}

Related MCP server: zentric-protocol-mcp

πŸ› οΈ Included Tools

Tool Name

Price (USDC)

Capability

prompt_injection_jailbreak_classifier

$0.045

High-speed deterministic classifier detecting indirect prompt injections, delimiter hijacking, and system role impersonation attacks in user inputs.

strip_prompt_injection

$0.02

Neutralizes indirect prompt injections, system prompt leak probes, jailbreak tokens, and hidden instruction tags in untrusted text.

secrets_entropy_scanner

$0.030

Shannon entropy and pattern analyzer scanning source code for leaked private keys, AWS access secrets, JWTs, and database connection strings.

obfuscate_pii_entities

$0.01

Detects and masks Personally Identifiable Information (SSNs, credit cards, emails, phone numbers, API keys) prior to model ingestion.

synthetic_dataset_bias_auditor

$0.040

Audits synthetic agent training data distributions for demographic bias, representation skew, and label drift across sensitive attribute categories.


πŸ”„ End-to-End 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.


πŸ’° The x402 Base L2 Micropayment Protocol

When an agent invokes a tool without payment, the server responds with a deterministic HTTP 402 Payment Required challenge containing:

  • Target tool price in USDC

  • Base Native USDC Contract: 0x833589fCD6eDb6E08f4c7C32D4f71b54bdA02913

  • Recipient payout wallet address

  • Single-use cryptographic nonce

Once broadcasted on Base L2, resubmitting with paymentSignature unlocks deterministic execution.


πŸ“„ License

MIT License. Created by Whambammy.

Available Tools

5 tools
obfuscate_pii_entitiesA

Detects and masks Personally Identifiable Information (SSNs, credit cards, emails, phone numbers, API keys) prior to model ingestion. (0.01 USDC on Base L2)

Inputs: payload (required), paymentSignature.

prompt_injection_jailbreak_classifierA

High-speed deterministic classifier detecting indirect prompt injections, delimiter hijacking, and system role impersonation attacks in user inputs. (0.045 USDC on Base L2)

Inputs: payload (required), paymentSignature.

secrets_entropy_scannerB

Shannon entropy and pattern analyzer scanning source code for leaked private keys, AWS access secrets, JWTs, and database connection strings. (0.030 USDC on Base L2)

Inputs: payload (required), paymentSignature.

strip_prompt_injectionA

Neutralizes indirect prompt injections, system prompt leak probes, jailbreak tokens, and hidden instruction tags in untrusted text. (0.02 USDC on Base L2)

Inputs: payload (required), paymentSignature.

synthetic_dataset_bias_auditorB

Audits synthetic agent training data distributions for demographic bias, representation skew, and label drift across sensitive attribute categories. (0.040 USDC on Base L2)

Inputs: payload (required), paymentSignature.

Tool Schema Changelog

Recent tool additions, removals, and schema changes observed during successful MCP inspections.

Loading tool schema changelog…

    TDQS

    A3.5/5.0

    Scored across 5 tools

    Disambiguation4/5

    Most tools target clearly distinct safety functions: PII masking, bias auditing, injection detection, injection neutralization, and secrets scanning. The two prompt-injection tools could be confused at a glance, but their descriptions distinguish detection/classification from stripping/neutralization.

    Naming Consistency3/5

    All names use snake_case and are descriptive, but the structural pattern is mixed: some are verb-first (obfuscate_pii_entities, strip_prompt_injection) while others are noun phrases ending in auditor/classifier/scanner. This is readable but not a predictable convention.

    Tool Count5/5

    Five tools is well within the sensible range for a focused AI safety/prompt shield server. Each tool covers a distinct guardrail capability, so none feels redundant or extraneous.

    Completeness4/5

    The surface covers key input-side guardrails: PII, bias, prompt injection detection/removal, and secrets scanning. Minor gaps remain around output-side moderation, toxicity classification, or policy enforcement, but the core lifecycle for a prompt safety server is reasonably covered.

    Related MCP Connectors

    Related MCP Servers

    • A
      license
      A
      quality
      B
      maintenance
      Protects AI agents from threats like prompt injection, jailbreaks, and SQL injection through a multi-layer scanning pipeline. It also enables PII redaction and rehydration to ensure data privacy during LLM interactions.
      12
      215 npm
      2
      Apache 2.0
    • F
      license
      Not graded
      quality
      D
      maintenance
      Security middleware for LLM apps and AI agent pipelines. Detects prompt injection attacks (22 signatures, 7 languages) and anonymizes PII (17 entity types). Deterministic, sub-25ms, GDPR Art.30 compliant.
      -
    • A
      license
      Not graded
      quality
      D
      maintenance
      Provides a pre-flight/post-flight firewall for LLM calls with comprehensive detection, classification, policy enforcement, reversible redaction, output safety, and immutable audit logging.
      1
      MIT
    • A
      license
      Not graded
      quality
      D
      maintenance
      Analyzes inputs and outputs in real-time to protect against prompt injections, data leaks, secrets exposure, and phishing URLs.
      22 npm
      3
      MIT