connections-optimizer
Pass
Audited by Gen Agent Trust Hub on Jul 29, 2026
Risk Level: SAFEPROMPT_INJECTION
Full Analysis
- [PROMPT_INJECTION]: The skill is susceptible to indirect prompt injection because it ingests untrusted data from external social media profiles during its inventory and scoring phases.
- Ingestion points: Following and connection inventories are retrieved from X and LinkedIn (SKILL.md).
- Boundary markers: The instructions lack explicit delimiters or 'ignore' instructions for the external profile data.
- Capability inventory: The skill can draft content for Apple Mail, X DMs, and LinkedIn, and use browser control for automation (SKILL.md).
- Sanitization: No specific filtering or sanitization of external profile content is described before processing.
- [SAFE]: The skill implements a safety-first approach by requiring a ranked action plan and drafts ('review pack') before any changes are applied, preventing autonomous pruning or messaging.
Audit Metadata