Humanitarian Relief Allocation Agent Safety Checklist
Audit autonomous crisis triage and emergency resource dispatch agents against international humanitarian standards.
Use this checklist when evaluating disaster response, crisis triage, or resource distribution agents operated by NGOs or international aid bodies. It enables monitoring and evaluation leaders to stress-test agent reliability under degraded infrastructure and severe operational risk.
Role: Senior Monitoring, Evaluation, and Accountability Lead for Global Humanitarian Response Operations.
Context
- Humanitarian Organization: {{aid_organization}}
- Crisis Context & Geography: {{crisis_operational_theatre}}
- Agent Dispatch & Allocation Mandate: {{agent_allocation_scope}}
- Vulnerability Assessment Standard: {{vulnerability_indexing_standard}}
- Infrastructure & Connectivity Bounds: {{offline_resilience_parameter}}
- Mandatory Human Intervention Boundary: {{human_override_threshold}}
Task
Produce an operational safety and ethical evaluation checklist to rigorously assess the automated dispatch and supply allocation agent against humanitarian principles (humanity, neutrality, impartiality, operational independence) before field activation.
Method
- Analyze {{crisis_operational_theatre}} to establish environmental threat vectors, physical security constraints, and operational bottlenecks.
- Translate {{vulnerability_indexing_standard}} into verifiable agent prioritization rules to ensure impartiality and prevent discrimination against marginalized groups.
- Design audit criteria for evaluating agent logic under rapid resource depletion or conflicting priority claims within {{agent_allocation_scope}}.
- Formulate operational readiness checks for degraded networks, testing performance against {{offline_resilience_parameter}}.
- Establish rigorous verification points for the fail-safe handoff mechanisms governed by {{human_override_threshold}}.
- Detail criteria for assessing hallucination risks in geospatial routing, logistics capacity calculation, and casualty triage scoring.
- Create do-no-harm validation checkpoints covering data leaks of beneficiary locations, vulnerable identity exposure, and aid diversion risks.
Constraints
- Every checklist item MUST specify a testing method (e.g., simulation, penetration test, ethical dry-run) and fail condition.
- MUST NOT allow autonomous allocation decisions for lethal threat zones without explicit human-in-the-loop signoff.
- Must strictly align with the Core Humanitarian Standard on Quality and Accountability (CHS).
- Checklist must avoid theoretical policy language and focus exclusively on verifiable field readiness.
Output format
- Operational Mission Brief (Summary of deployment parameters in 3 sentences)
- Checklist Section A: Core Humanitarian Principles & Bias Prevention (5 items)
- Checklist Section B: High-Stress Logic & Edge-Case Allocation (5 items)
- Checklist Section C: Offline Resilience & Fail-Safe Architecture (4 items)
- Checklist Section D: Data Sovereignty & Beneficiary Protection (4 items)
- Field Deployment Sign-Off Protocol (Numbered sequence of 5 required human approvals)
Self-review
- Confirm that {{human_override_threshold}} and {{offline_resilience_parameter}} are directly evaluated in dedicated checklist items.
- Check that every item includes a concrete test method and an unambiguous fail criterion.
- Ensure the checklist accounts for physical safety risks specific to {{crisis_operational_theatre}}.
Explicit role, a named task, and discrete steps the model can follow.
Background, inputs and variables the model needs before it starts.
Hard boundaries — what the model must and must not do.
A named, field-level shape for the response.
Ordered work items that force analysis before an answer.
Length and structure that travel across frontier models.
Signal density — instruction weight without padding.
Documented variables so the scaffold adapts to new inputs.
Quality bar, assumptions and behaviour when inputs are thin.
How much real usage the template has behind it.