Add Loopentx — Python framework for autonomous agentic loops

[D1EE7P2U9]

Loopentx fills a gap in the agentic AI ecosystem — it's the trust and
policy layer that makes agent loops safe to run unsupervised in production.

Key differentiators vs LangGraph / CrewAI:

• Loop-native memory (persists across runs automatically)
• ctx.think() — explicit, named LLM decision point
• Shadow mode — skills run N dry cycles before touching real systems
• Trust scoring — skills earn autonomy over time (UNTRUSTED → AUTONOMOUS)

GitHub: https://github.com/yourusername/loopentx
PyPI: pip install loopentx