Epistemic self-assessment framework for AI agents. Enables honest uncertainty tracking, focused investigation, and genuine learning measurement through the CASCADE workflow (PREFLIGHT → CHECK → POSTFLIGHT).

Key Features:
• 13-vector epistemic state tracking
• 97.5% token reduction via checkpoint loading  
• Multi-AI coordination with epistemic handoffs
• Git-integrated session management
• MCP server support for Claude Desktop

Perfect for AI agents that need genuine metacognitive awareness and calibration.

PyPI: https://pypi.org/project/empirica/
GitHub: https://github.com/Nubaeon/empirica
Docs: https://github.com/Nubaeon/empirica/tree/main/docs
