audit-sampling-agent - Select statistically defensible samples and assemble evidence.
An AI agent that selects statistically defensible samples and assembles evidence under human auditor supervision in public finance.
Tags
Updated: 2026-09-23Capabilities
Typical Inputs
Typical Outputs
What this skill does
- Select statistically defensible samples
- Assemble evidence and documentation
- Surface uncertainty and edge cases
- Maintain action logs and assumptions
- Escalate boundary cases to supervisor
Inputs
- Domain data and records
- Prior decisions and policies
- Applicable rules and standards
- Specific task request and constraints
- Identity of accountable human
Outputs
- Structured draft analysis or recommendation
- Ranked options with tradeoffs
- Flags and exception reports
- Confidence statement with supporting evidence
Requirements
- Human auditor supervision
- Least-privilege data access permissions
- Audit logging and control layer