Docs · hypothesis-led

Company memory

Memory updates are proposed lessons, not automatic chat history. Only approved, retrievable memory counts as compounding. Unapproved interpretation must not silently become product truth.

Agent memory boundary Try a mission

Docs hub Get started llms.txt Agent capabilities

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

Lifecycle

Propose → approve or reject → retrieve on later missions → prove use when it influences a run.

Anti-pattern

Storing text that is never retrieved, never approved, or never bounded by claim ceilings.

Next safe action

Learn the contract, then run a bounded mission.

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

Quickstart Agents