prioritizing-my-inbox
Pass
Audited by Gen Agent Trust Hub on Nov 5, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: The skill is a read-only utility designed to rank and display internal reports. It uses the platform's
activity-summary-getandinbox-reports-listtools to fetch context and content. Findings show no signs of prompt injection, data exfiltration, or malicious execution. - [SAFE]: Data processing follows a strict scoring logic defined in
references/scoring.md. This logic incorporates priority levels, model predictions (lifts), and user activity without introducing external or untrusted data sources. - [SAFE]: The skill includes explicit guardrails to prevent unauthorized actions, such as claiming or dismissing reports, and instructs the agent to hand off to
inbox-explorationfor write operations. - [SAFE]: References to external entities (like GitHub PRs) are part of the standard PostHog integration and do not involve suspicious downloads or remote code execution.
Audit Metadata