Module agent_policy 

What an agent is in a room, decided in one place.

An agent is an ordinary collaborator acting for one person: it never holds more than that person’s role, never more than what its room admission grants it, and never more than [ROLE_CAP]. Every host — the relay governing a room, every member admitting its ops, the desktop broker hosting a room itself — asks [role] rather than deciding on its own.

Four questions about agents are still open. Until they are decided each keeps the answer the first agent-collaborator release gave it, and each is changed here:

1. Is an agent’s role capped at Editor? [ROLE_CAP]: yes.
2. May another person’s agent join a room? [OTHER_PEOPLES_AGENTS_JOIN]: no; a room’s governor admits only its own person’s agents.
3. What does withdrawing an agent mean? [WITHDRAWAL_IS_FINAL]: a withdrawn agent device is never admitted again; granting it anew needs a new agent key.
4. Do members enforce the app scope of an agent’s delegation? Not yet: the agent’s host does, and members enforce the delegation’s rooms and expiry only. Enforcing it at members needs the app on every op, which the room’s ops do not carry.

Constants

OTHER_PEOPLES_AGENTS_JOIN

Whether a room’s governor admits an agent of a person other than its own (open question 2; see the module documentation).

ROLE_CAP

The most an agent ever acts as, whoever it acts for. An agent never governs a room.

WITHDRAWAL_IS_FINAL

Whether withdrawing an agent from a room is final for its device (open question 3; see the module documentation).

Functions

role

The role an agent acts with when its person holds person and its room admission (its preset, or its delegation) grants it granted: the least of the two and ROLE_CAP. Never more than either.