Agent♥︎Age
Catalog

Lovec

Official

by lovec-tech · Python

Prompt-injection detector for untrusted text before it reaches an LLM (RU/EN). BYO lovec.tech key.

This MCP server detects prompt injection in untrusted text before it reaches an LLM. It supports RU/EN input and can use a lovec.tech API key provided by the user (“BYO”). The server is intended to help evaluate whether text may contain injection attempts.

🛠️ Key Features

  • Prompt-injection detection for untrusted text before LLM processing
  • RU/EN coverage
  • Requires a user-provided lovec.tech key

🚀 Use Cases

  • Pre-processing untrusted prompts to reduce prompt-injection risk
  • Handling mixed RU/EN text from external or user-provided sources

⚡ Developer Benefits

  • Clear separation between untrusted input and LLM ingestion
  • Fits into MCP-based workflows for LLM security checks

⚠️ Limitations

  • Depends on a lovec.tech key supplied by the user (“BYO lovec.tech key”)

Topics

ai-safetyllm-securitymcpprompt-injection

Related servers

More in Security