Skip to main content
Glama
genpark-agent-prompt-injection-sanitizer-firewall-skill

genpark-agent-prompt-injection-sanitizer-firewall-skill

genpark-agent-prompt-injection-sanitizer-firewall-skill

Python 3.9+ License MIT MCP Compatible GenPark AI Zero Dependencies

🌐 GenPark MCP Hub β€’ πŸ“¦ GenPark Official β€’ πŸ“– Documentation


πŸ“Œ Overview & Capability

genpark-agent-prompt-injection-sanitizer-firewall-skill is a deterministic, high-performance, zero-dependency Python tool and native Model Context Protocol (MCP) server designed for autonomous AI agents, multi-agent frameworks (LangGraph, CrewAI, AutoGen, OpenAI Swarm), and developer environments (Cursor, Windsurf, Claude Desktop).

Executive Capability: Heuristic and pattern-based agent prompt injection sanitizer and LLM jailbreak firewall detecting adversarial jailbreaks, system prompt exfiltration, and tool hijacking.

⚑ Key Highlights

  • 🐍 Zero External pip Dependencies: Implemented entirely with pure Python standard library for instant zero-overhead execution.

  • πŸ”Œ Native Model Context Protocol (MCP): Plugs directly into any MCP-compliant client via JSON-RPC 2.0 stdio.

  • ⚑ Sub-Millisecond Execution: Slashes token burn and latency by resolving routine agent tasks deterministically without frontier LLM round-trips.

  • πŸ›‘οΈ Production-Hardened: Comprehensive error handling, boundary validation, and telemetry.


Related MCP server: aletheia-mcp

πŸ—οΈ Architecture

graph LR
    Agent([πŸ€– Autonomous Agent / IDE]) -->|MCP Protocol / JSON-RPC| Server[⚑ genpark-agent-prompt-injection-sanitizer-firewall-skill Server]
    Server --> Core[🧠 Deterministic Processing Core]
    Core --> Out[πŸ“Š Actionable Result & Telemetry]
    Out --> Agent

πŸš€ Quickstart & Usage

1. Direct Python Client Execution

python example_usage.py

2. Programmatic Integration

from client import AgentPromptInjectionSanitizer

client = AgentPromptInjectionSanitizer()
result = client.run_firewall_benchmark()
print(result)

πŸ”Œ Model Context Protocol (MCP) Setup

Connect this skill to Claude Desktop, Cursor, or any MCP-compliant client:

claude_desktop_config.json

{
  "mcpServers": {
    "genpark-agent-prompt-injection-sanitizer-firewall-skill": {
      "command": "python",
      "args": ["/path/to/genpark-agent-prompt-injection-sanitizer-firewall-skill/mcp_server.py"]
    }
  }
}

Direct MCP Testing

python mcp_server.py --test

πŸ“Š Technical Specifications

Parameter

Type

Required

Description

payload

string / dict

Yes

Primary context, code, schema, or content input

options

dict

No

Execution flags, compression ratios, or risk bounds


Tool Schema Changelog

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

Loading tool schema changelog…

    Related MCP Connectors

    Related MCP Servers

    • A
      license
      Not graded
      quality
      D
      maintenance
      A defensive gateway and firewall for AI agents using MCP servers, scanning tool calls, responses, and manifests for prompt injection, secrets, dangerous commands, and drift before allowing execution.
      MIT
    • A
      license
      A
      quality
      B
      maintenance
      Enables sub-millisecond pre-execution safety filtering for AI agent tool calls, blocking destructive shell commands, dangerous SQL mutations, credential access, SSRF, and scope-creep or prompt-injection patterns before they execute, with optional fail-closed transparent proxy wrapping for any MCP server.
      6
      58 npm
      12
      Apache 2.0
    • F
      license
      Not graded
      quality
      B
      maintenance
      Enables deterministic detection and neutralization of adversarial prompt injections and override attempts in AI agent workflows via a zero-dependency MCP server, providing structured telemetry and low-latency validation.
      8
      -
    • A
      license
      Not graded
      quality
      C
      maintenance
      Provides a three-layer firewall for MCP servers that inspects tool outputs for malicious content, enforces egress destination policies, and gates consequential tool calls. It runs as both a proxy and an MCP server, enabling agents to scan content, check tool calls, and evaluate results before they reach the model.
      57 npm
      1
      MIT