proof-writer

Draft and verify mathematical proofs for ML/AI claims with explicit assumptions and dependency maps.

Updated May 29, 2026
One-click install
npx skills add https://github.com/Mang30/myskills --skill proof-writer-mang30
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: proof-writer
Source: https://github.com/Mang30/myskills/tree/main/skills/proof-writer
Command: npx skills add https://github.com/Mang30/myskills --skill proof-writer-mang30

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

It prevents incorrect or hand-wavy reasoning by producing a rigorous, assumption-aware proof (or a clear blockage report) for theorems and lemmas in ML/AI theory.

Core Features & Use Cases

  • Rigorous proof drafting: Writes honest proofs for theorems, lemmas, propositions, or corollaries, including quantifier- and boundary-case care.
  • Proof feasibility triage: Classifies claims as provable as stated, provable only after weakening/extra assumptions, or not currently justified.
  • Structured proof packages: Outputs a proof document with explicit claim, assumptions, notation, proof status, dependency map, and verification.

Quick Start

Ask the AI to “prove the following theorem under the stated assumptions, and write the result into PROOF_PACKAGE.md with a proof status and dependency map.”

Frequently Asked Questions about proof-writer

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I write a rigorous mathematical proof for an ML theory claim?

To write a rigorous mathematical proof, extract the exact statement and assumptions, triage feasibility without fabrication, and draft an explicit, verification-ready proof document. This ensures assumption-aware rigor for theorems and lemmas in ML/AI theory.

What is proof feasibility triage and when do I need it for theorem proving?

Proof feasibility triage classifies claims as provable as stated, provable only after weakening or adding extra assumptions, or not currently justified. You need it when determining whether a mathematical argument can be completed before drafting a full proof.

How to fill missing steps in a mathematical proof sketch?

To fill missing steps in a proof sketch, apply rigorous proof drafting by extracting the exact statement, assumptions, and notation. This verifies the sketched argument and completes the missing steps with quantifier- and boundary-case care.

Can I verify if a claimed ML theorem argument can be completed without making up steps?

You can verify a claimed argument by applying proof feasibility triage, which evaluates whether the theorem is provable as stated or requires extra assumptions, triaging feasibility without fabrication and blocking incomplete attempts.

What should a structured proof package include for verification?

A structured proof package for verification should include the explicit claim, assumptions, notation, proof status, dependency map, and the verification-ready proof document itself, ensuring all boundary cases and quantifiers are properly handled.

What are the limitations of using automated proof writing for ML theory?

Automated proof writing for ML theory limits outputs to provable claims or clear blockage reports, meaning it cannot fabricate steps for unprovable arguments and will block attempts that require weakening assumptions or extra conditions to complete.