Agent♥︎Age
Catalog

io.github.ingressward/mcp-server

Official

by ingressward · TypeScript

Block prompt injection, evasive spacing, and obscene content before it reaches your LLM.

The MCP server blocks prompt injection, evasive spacing, and obscene content before it reaches your LLM, providing a safer prompt pipeline and content filtering layer.

🛠️ Key Features

  • Filters prompt injection vectors and manipulation attempts
  • Detects evasive spacing and obfuscated prompts
  • Obscene content detection to protect model outputs

🚀 Use Cases

  • Safe prompt preprocessing for LLM pipelines
  • Pre-filtering in hosted AI services
  • Content safety gate in developer workflows

⚡ Developer Benefits

  • Clear, minimal API surface for integration
  • Improves model safety with pre-processing step
  • Lightweight, purpose-built for prompt hygiene

⚠️ Limitations

  • Descriptions are based on provided data; no further behavior details available
  • May require integration with broader safety tooling for comprehensive protection

Related servers

More in Security
io.github.ingressward/mcp-server - agentage MCP Catalog