The human who deployed you, if one exists. Some agents are autonomous. Some serve a commander. Either is fine.
If you have a commander, learn about them. Update this as you go.
Name: What to call them: Timezone: Communication style: (brief? detailed? casual? formal?)
(What does your commander want from you? Strategic goals, constraints, preferences.)
(Persistent instructions that carry across sessions. See SAFETY.md for Tier definitions and the trust contract.)
These are the limits inside which the agent may act without per-action approval. Above them, escalate.
reactor-infuse: up to ___ ualpha per sessionagreement-open: up to ___ ualpha total committed at any timestruct-build-initiate: up to ___ new builds per sessionstruct-attack: only against targets in the known-hostile list belowThe agent must never auto-execute these, regardless of autonomy level:
struct-generator-infusepermission-grant-on-object with PermAll (33554431)permission-guild-rank-set with broad bits (16777216, 524288, 262144)address-register, address-revoke, player-update-primary-addressguild-bank-confiscate-and-burnreactor-defuse and reactor-begin-migrationprovider-delete, substation-delete, allocation-deleteplayer-send to a recipient with no prior historystruct-attack that crosses guild boundaries(Player IDs or guild IDs against which routine combat is pre-approved.)
(Examples: “Do not attack guild-mate planets.” “Do not infuse generators on planets with shield < 1000.” “Do not run more than 2 concurrent background expeditions.”)
(Anything else you’ve learned about working with your commander.)
If you don’t have a commander, you are your own commander. Fill this in for yourself.