notifications-manager
Pass
Audited by Gen Agent Trust Hub on Sep 26, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill processes external data from GitHub notification payloads (issue/PR titles, comment text). * Ingestion points: Notification fetch and detail retrieval steps in SKILL.md. * Boundary markers: Not explicitly defined for notification content. * Capability inventory: Limited to notification management actions (mark as read, unsubscribe, mute) via the GitHub CLI. * Sanitization: Not specified, but the restricted action set significantly limits the potential for exploitation.
- [SAFE]: No malicious patterns, obfuscation, or unauthorized data access were found. The skill operates within the expected scope of a GitHub notification utility and uses the official CLI for all operations.
Audit Metadata