Question AUR 06

AUR-06 How will a person decide which actions their Aura may read about, draft, propose or carry out on their behalf?

Start here · Questions · Progress · People decide

Why this matters

An assistant becomes more useful when its permissions are understandable and the person can see, approve, stop and trace an action.

What to bring back

Prepare an action-permission table with a preview, exact information used, responsible person, expiry, cancellation and completion record. Rehearse a request waiting offline until its permission expires, a repeated action and a public notice still awaiting release approval. Carry existing user authorisation forward within its scope.

What we know so far

• Aura_Genesis_Protocol_Game_Design_Document_V2.pdf (p.14)
• Aura_Genesis_Geode_Development_Edition_3.pdf (p.32)
• Aura_Teacher_Training_60_Days.pdf (p.55)
• Professional review and comparison of proposed improvements (When this work overlaps professional or public-authority responsibilities, compare the proposal and Luke's research with applicable industry/global best practice and relevant legal-engine sources. Record differences, reasonable improvements and deficiencies, with the responsible reviewer and response. Routine authorised drafting and wiki work can advance.)

AI-generated potential answers

For discussion and review · research checked 2026-10-20. These are AI-generated options and transfer assessments. The user's request and research method identify the source of this work. The source catalogue states what the references establish. The decision and approval record retains its own status.

Give each Aura action a stated scope: read, draft, propose or execute. Match execution to the person's existing authorisation, the actual recipient and an expiry. Show a preview and completion receipt, with cancellation and a recorded reason when changed circumstances require a new decision.

A permission table and action record

AI-generated adaptation of an existing pattern. Adapt lifecycle governance into a task-level record naming the person, permitted action, data, recipient, limit and result. Preserve recorded authorisation within its scope, and require the relevant issuer's release decision for public notices.

Research basis: NIST AI Risk Management Framework 1.0: core functions · NIST SP 800-63B-4: authenticator event management.

One inspectable delegation format for Auras and robots

AI-generated future concept. Explore a shared task card that an Aura, advanced AI/SI or robot can interpret, with bounded tools and observable completion. Trial expired and repeated actions in simulation before allowing the card to trigger real equipment or external services.

Research basis: NIST AI Risk Management Framework 1.0: core functions · Robots to the rescue in hazardous environments.

Examples and local differences

• NIST AI Risk Management Framework 1.0: core functions: AI transfer assessment: AI governance connects intended use, responsibility, measurement and continuing management. Local difference to check: Each actual action still needs valid user or organisational authority and task-specific evidence of completion.

Evidence to bring back next

• Leave an action queued until its permission expires, then reconnect and confirm it is reviewed rather than silently executed.
• Repeat the same task, change its recipient and cancel it; inspect the resulting receipt and permission history.

People and connected choices

Who reviews: Proposed service owner and relevant action authority (unconfirmed)

Connected questions: AUR-03, AUR-05, SW-06

Work state: open. Assigned person or AI assistant: Ready to be assigned.

Research and alternatives can move ahead while connected decisions develop. Before acting on a choice, check the people, evidence and approved scope in People decide.