The tool that won't let AI say anything it can't cite

Build an open-source library that forces citation-heavy responses from any LLM using a verifiable chain of thought. Ensure the model output matches retrieved context fragments.

Suggested repo: citeForce

"If the model can't cite it, it can't say it."

Estimated effort: 100h