ProofDama: Certifying Italian Checkers

Note: This analysis is based entirely on the self-reported project description provided by the author. No external verification, revenue data, customer base, or traction evidence is available.

The description states that ProofDama is a proof-engineering platform for solving Italian Checkers under official rules. It separates three concepts:

• A program that plays strongly;
• A search process that reports a result;
• An independently replayable mathematical certificate.

Only the third constitutes proof.

It includes:

• A canonical Rust implementation of Italian rules;
• Graph-aware, resumable depth-first proof-number search;
• Deterministic parallel graph construction;
• Ed Gilbert WLD and MTC endgame evidence;
• Content-addressed artifacts with SHA-256 manifests;
• Transactional coordinator for multiple workers;
• Atomic checkpoints that survive interruption;
• Independent verifiers that regenerate every legal move instead of trusting the solver.

The system is described as hybrid, using both Apple Silicon and Linux servers, with GPT-5.6 in Codex assisting in architecture design, code review, and verification tasks.

Claim: The product is a formal proof system for Italian Checkers.

Evidence: Self-reported by author; no independent validation provided.

The author positions ProofDama as:

• A continuous research campaign toward solving the standard 8×8 game;
• Not merely a strong-playing engine, but a platform for producing mathematical proofs;
• An open and reproducible system that avoids heuristic evaluations in favor of certified results.

It distinguishes itself from prior work by:

• Using deterministic parallelism to ensure reproducibility;
• Requiring independent replay of all results;
• Rejecting unverifiable claims like minimax or heuristics.

The project evolved from a personal academic exercise into a formal scientific endeavor, enabled by GPT-5.6 and Codex.

Claim: ProofDama aims to produce weak solutions for Italian Checkers.

Evidence: Self-reported; no external confirmation of success or failure.

Not evidenced.

The description does not identify any specific customer segment, target user group, or buyer persona. It focuses on the scientific and technical aspects of solving a game rather than commercial adoption or usage.

Finding: No evidence of defined target customers or ideal customer profile (ICP).

Not evidenced.

There is no mention of pricing models, monetization strategies, or business structures in the description. The project appears to be a research effort without any indication of commercial intent or revenue generation.

Finding: No evidence of business model or pricing information.

The system uses:

• Rust for canonical rules and proof core;
• Python for coordination and scripting;
• TypeScript/Next.js for public-facing website;
• GPT-5.6 in Codex for engineering assistance;
• Distributed computing across Apple Silicon and Linux machines;
• SHA-256 manifests, content-addressed artifacts, atomic checkpoints;
• Independent verification of all states and transitions.

It distinguishes between:

• WLD (win/draw/loss) evidence;
• MTC (move-count) linked forward evidence;
• Fully certified results.

The architecture supports:

• Resumable jobs;
• Deterministic parallelism;
• Cycle handling in game graphs;
• Cross-platform reproducibility.

Claim: ProofDama is technically robust and reproducible.

Evidence: Self-reported; no independent audit or performance metrics provided.

Not evidenced.

There is no evidence of:

• Revenue;
• Customers;
• Adoption;
• Product usage;
• Market traction.

The project is described as ongoing, with only completed reduced-board proofs (4×4×1, 4×6×1, 4×8×1) shown. The full 8×8 game remains unsolved.

Finding: No evidence of traction or maturity beyond initial proof-of-concept results.

Not evidenced.

No mention of competitors, market dynamics, or competitive positioning is included in the description. The focus is purely on technical execution and scientific rigor.

Finding: No evidence of competitive landscape or market context.

Several risks are implied by the self-report:

1. Trust in GPT-5.6: The author relies heavily on Codex for system design, code review, and verification. If these tools fail or produce incorrect outputs, the entire proof may be compromised.
2. Hardware dependency: The system depends on specific hardware (Apple Silicon, Hetzner servers), which introduces risk of failure or inconsistency if not maintained properly.
3. Reproducibility assumptions: While the system claims deterministic parallelism and byte-level reproducibility, this has not been independently verified.
4. Scalability concerns: The project faces state-space explosion issues, and there is no evidence that it can scale to full 8×8 boards without significant resource constraints or architectural changes.

Inference: The reliance on AI tools and hardware-specific execution raises questions about long-term reliability and generalizability.

1. What are the exact mechanisms by which GPT-5.6 was used in generating or verifying proofs? Is there a documented audit trail?
2. How is the independence of verification ensured across different platforms (ARM vs x86)?
3. Has the system been tested under adversarial conditions or with corrupted data to validate its robustness?
4. What are the actual resource requirements for solving larger boards, and how does this impact scalability?
5. Are there any known limitations in the current implementation that could prevent full 8×8 solution?
6. How is the integrity of the proof certificates validated beyond SHA-256 checksums?

Note: These questions aim to probe the verifiability and reproducibility claims made in the description.