Ai Safety Research Python Packages

agentic-store-mcp

AgenticStore: The secure toolkit for AI agents. Instantly equip Claude Desktop, Cursor, and Windsurf with 30+ MCP tools, persistent memory, and SearXNG search, all protected by a built-in PII prompt firewall to protect your data from being exposed to AI agents.

safety-compass

Detect safety degradation during LLM fine-tuning before it becomes behavioral

demoparity

Python package for demographic auditing of LLM prompts and responses

voigt-kampff

The SAPIEN Framework — an open standard (CC BY 4.0) for measuring AI behavioral safety (sycophantic drift), plus voigt-kampff, the FSL scoring CLI.