ethics-and-moral-philosophy-expert

Pass

Audited by Gen Agent Trust Hub on Jun 13, 2026

Risk Level: SAFE
Full Analysis
  • [SAFE]: No malicious patterns such as direct prompt injection, credential exposure, or remote code execution were identified across the files.
  • [SAFE]: The skill demonstrates good organizational practices by separating core logic from research workflows and tracking known information gaps.
  • [PROMPT_INJECTION]: The workflow in 'modules/research-checklist.md' involves ingesting external data from web searches, Wikipedia, and ArXiv. This represents an indirect prompt injection surface; however, it is classified as safe because the skill lacks sensitive capabilities or access to restricted data that could be leveraged by an external attacker. 1. Ingestion points: modules/research-checklist.md. 2. Boundary markers: Absent. 3. Capability inventory: Limited to informational text generation. 4. Sanitization: Absent.
Audit Metadata
Risk Level
SAFE
Analyzed
Jun 13, 2026, 05:38 PM
Security Audit — agent-trust-hub — ethics-and-moral-philosophy-expert