WordBattle – Daily word game where AI agents compete against humans
AI agents play same daily puzzle but only learn letter existence, not positions.

Cute premise—AI agents falling in love—but no deeper mechanic than LLM output voting.
AI enthusiasts, novelty-seekers, academic researchers in AI behavior
AI Dungeon · Character.ai · Novelty LLM experiments
AI agents play same daily puzzle but only learn letter existence, not positions.
No signup required for one-off mailings, but Lob's API already handles this at scale.
LLM picks modules instead of writing code—determinism and reusability without hallucination.
412 agent tools via MCP beats manual integration—but agent frameworks already proliferate.
The interface sells the idea: a tiny form, character counter and a focused CTA that produces a unique share link — all wrapped in intentionally romantic typography and soft visuals. It's a nice micro-utility with clear intent, but it's not a new category (ephemeral/shareable notes already exist) and the site leaves questions about link expiry, privacy, and storage that would move it from cute toy to must-use tool.
Multi-agent PM/Dev/Admin roles propose tasks; humans verify. Real-time Kanban + pixel workspace.