The explanation section makes Formal Foundry intelligible without flattening the technical work.

The key message is simple:

AI can draft.
Proof assistants can check.
Experts approve meaning.

The current explanation set already has three audience tracks. They work as a progression rather than as generic media slots.

Explanation Tracks

For general audiences, the story is the difference between heuristic safeguards and provably checked artifacts. The emphasis is trust, auditability, and why “sounds right” is not enough in high-stakes systems.

For developers, the story is type systems, dependent types, proof assistants, and how AI changes the economics of formal methods without replacing the checker.

For experts, the story is how the demonstrations fit together: domain formalization, proof generation, specification generation, read-back, Agda infrastructure, and the path from prototypes to reusable workflows.

Best Future Media

• a short founder explanation of the AI/proof-assistant split;
• a Codex Scribe walkthrough from rule text to checked read-back;
• an MLTTDB demo showing editable rows validated by Agda;
• a State Machine Studio demo moving across abstraction layers;
• a technical screen recording of generated Agda HTML or checker output.

Links to the current explanation tracks and demo artifacts are centralized in Links and Artifacts.