linkedin-interviewer
Pass
Audited by Gen Agent Trust Hub on Sep 16, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: No malicious patterns or security vulnerabilities were detected. The skill focuses on structured data collection for professional writing.
- [INDIRECT_PROMPT_INJECTION]: The skill identifies a surface for indirect prompt injection when processing user-provided text or external data (e.g., from Apify). It mitigates this risk with specific instructions: 1. Ingestion points: User chat input, external text pastes, and files like references/voice-profile.md and story-bank.md. 2. Boundary markers: Explicitly directs the agent to treat such content as data rather than instructions and references a dedicated untrusted-content.md guideline. 3. Capability inventory: Limited to reading and writing local material files; no network access or shell execution is performed (SKILL.md). 4. Sanitization: Instructs the agent to prioritize user conversation over pasted facts and to verify naming/confidentiality limits directly with the user.
- [DATA_EXPOSURE]: While the skill collects sensitive professional details (scars, numbers, NDAs), it implements protective measures by advising the user to gitignore the data file and requiring explicit permission for naming entities in drafts.
Audit Metadata