proof-writer

To write rigorous mathematical proofs for ML/AI theory, you provide the exact theorem statement, assumptions, and any proof sketch to generate a structured, verification-ready PROOF_PACKAGE.md with stepwise justifications.

Formalizing an informal ML theorem involves normalizing the claim, building a dependency map, and drafting a structured proof package that explicitly states assumptions and ensures logical validity for formal verification.

Structuring a proof package requires organizing the output into sections for Claim, Assumptions, Notation, Proof Strategy, Dependency Map, and Stepwise Justification to ensure logical validity and rigorous justification.

Yes, if your proof sketch fails, the workflow can output a corrected claim with a complete proof, or it will generate a detailed blockage report explaining exactly why the claim cannot be justified.

When theorem proving is blocked by missing assumptions, the process outputs a detailed blockage report identifying the logical gaps, rather than forcing an invalid proof, ensuring formal correctness.

No, formal proof writing for ML theory does not require a complete strategy upfront; you simply provide the theorem statement and any available assumptions or proof sketches to initiate the workflow.