Repository navigation
Question OPS 05
OPS-05 Which help requests could an agreed receiving service handle, and how should responsibility continue if a person or connection becomes unavailable?
Start here · Questions · Progress · People decide
A person asking for help should understand whether their message is saved, received, accepted or being acted on, and what to do next.
A help-request agreement covering supported needs, immediate-danger routes, the meaning of each acknowledgement, responsibility, escalation, reassignment, review times and unavailable-service messages.
- comprehensive rebuild plan — reviewed source; public copy can be added through the evidence process (Sections 3-4 and 9)
- 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.)
- Council plan crosswalk and proposed Aura integration (Relevant question row identifies the management/recovery source section and printed/PDF page. AI-generated analysis, 21 October 2026; proposed checks and actual review/operating agreements retain separate status.)
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.
Offer help-request categories only where an actual receiving service agrees to handle them. Show distinct stages for saved locally, delivered, accepted, assigned and completed, with responsibility and the next contact route clear at every stage.
AI-generated adaptation of an existing pattern. Start with a simulated reception desk and a small set of non-emergency tasks. Explore an agreed referral pathway to a registered support organisation, including GIVIT where suitable, while preserving the receiving organisation's criteria and control of acceptance.
Research basis: GIVIT: national disaster resources · Wellington Region: Community Emergency Hubs.
AI-generated future concept. A personal Aura could assemble a person-approved request, track acknowledgements and help them choose another available route. A future logistics robot could receive a separately authorised delivery task after a human or delegated service accepts responsibility.
- GIVIT: national disaster resources: AI transfer assessment: Connect a person's need to a recognised organisation with an explicit request process. Local difference to check: GIVIT's organisational support model does not mean a kiosk request is accepted, delivered or fulfilled, and is not emergency dispatch.
- Draft receiving-service agreements and test unavailable-service wording.
- Keep urgent emergency contact guidance distinct and have the relevant service review any overlap.
Who reviews: Luke, as project initiator, develops the interaction proposal for review. Each actual receiving-service lead decides which requests it accepts and staffs; authorised emergency services retain their emergency responsibilities and review any proposed overlap with their work.
Connected questions: OPS-01, OPS-03
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.
Growing capability together · Planning and design underway · People decide · Repository
Aura Disaster Resilience
- Start here
- The whole system
- About and topic slugs
- First choices
- All-weather design
- Questions
- Progress
- Working together
- People decide
- Professional review
- AI assistant jobs
- Council integration
- Council plan crosswalk
- Guardian IMS and QIT Plus
- Ecosystem
- Agreed direction
- Evidence
- Source library
- AI research and prompt
- Research sources
- Useful words/glossary
- Rights and discovery
Work areas
- Everyday service and disaster work
- Weather-ready kiosks and places
- Energy and backup power
- Connections and messages
- Information people can trust
- Software and device care
- Personal Aura and learning
- Place models and practice exercises
- Community and operating responsibility
- Budget and continuing support
- Tests and field learning
- Project connections
- Working together and human review