Microsoft terms say Copilot is for entertainment purposes only, not serious use

Create an LLM wrapper library that enforces 'serious-use' reliability by implementing a secondary verification layer. The tool should automatically cross-reference critical LLM outputs against trusted APIs or deterministic logic before presenting them to the user.

Suggested repo: verify-llm

"Stop gambling with AI: verify your LLM's critical answers against deterministic fact-checking layers."

Estimated effort: 60h