Docs · hypothesis-led

Receipts

A receipt is durable, inspectable proof that work ran under a scope. Terminal state and side-effect status are never inferred from UI polish. Model output alone is not a completion receipt.

Inspect a sample path Agent contract

Docs hub Get started llms.txt Agent capabilities

Hypothesis-led documentation. No ranking, traffic, or outcome guarantees. Public discovery is not execution authority.

Minimum truth

Actor, scope, terminal state, side-effect status, and evidence refs for completed work. Trace correlation when available.

Receipt classes

Run, gate, side-effect, and memory-approval receipts are different objects. Collapsing them into one green badge creates false confidence.

Marketing twin

See /agent-receipts for the founder explainer.

Next safe action

Learn the contract, then run a bounded mission.

Prefer receipts over claims. Prefer fail-closed honesty over impressive incompleteness.

Quickstart Agents