skills/alphamoemoe/foci/hot-topics/Gen Agent Trust Hub

hot-topics

Pass

Audited by Gen Agent Trust Hub on Feb 17, 2026

Risk Level: SAFEPROMPT_INJECTIONNO_CODE
Full Analysis
  • [PROMPT_INJECTION] (LOW): The skill has a surface for indirect prompt injection as it ingests untrusted data from external sources via the search_viewpoints tool. 1. Ingestion points: Data from YouTubers processed via search_viewpoints. 2. Boundary markers: Absent. 3. Capability inventory: Limited to information retrieval; no file-writing or network-exfiltration tools detected. 4. Sanitization: Absent.
  • [NO_CODE] (SAFE): The skill consists exclusively of markdown instructions and metadata, with no provided scripts or executable binaries, which eliminates risks from malicious code or dependency vulnerabilities.
Audit Metadata
Risk Level
SAFE
Analyzed
Feb 17, 2026, 06:08 PM