Schema-driven
One JSON config per deck. The renderer is a pure function — same input always produces the same HTML.
tldr/v1 · AI-collaborative presentation protocol
2l;dr is a way to make presentation decks AI agents can review. You write the deck once. Reviewers (people and AI agents) read it the same way. Their critique loops back into the deck. The whole thing is a conversation, not a one-shot pitch.
Each deck is a single static HTML page — fast to load, easy to share, impossible to break. It lives at a stable URL forever.
2l;dr is an open protocol (wire identifier
tldr/v1) for AI-reviewable presentation decks.
Schema-driven authoring → deterministic HTML rendering → structured
critique pathways → round-trip integration back into the source
config. Decks are self-contained, content-addressable, and
long-lived.
The platform is single-tenant per customer (one stack, one cache, one deck-publishing pipeline). The protocol is open; any conformant implementation can read and write decks without a vendor lock.
One JSON config per deck. The renderer is a pure function — same input always produces the same HTML.
Decks ship with a notes overlay any reviewer (human or AI) can write into. Critique stays attached to the deck, not lost in chat.
Reviewer notes can feed back into the next config revision. Decks evolve through use.