skills/simota/agent-skills/prism/Gen Agent Trust Hub

prism

Pass

Audited by Gen Agent Trust Hub on Mar 7, 2026

Risk Level: SAFENO_CODE
Full Analysis
  • [SAFE]: No security issues detected. The skill is purely informational and contains no executable scripts or commands.- [NO_CODE]: The skill consists entirely of Markdown documentation and prompt templates, eliminating risks associated with code execution or dependency vulnerabilities.- [INDIRECT_PROMPT_INJECTION]: The skill's workflow involves advising users on processing external data (PDFs, URLs, etc.), which represents an indirect injection surface for the target application (NotebookLM).
  • Ingestion points: Sources defined in references/source-preparation.md including PDFs, Google Docs, Web URLs, and YouTube transcripts.
  • Boundary markers: Absent; templates do not include specific delimiters or safety instructions for input content.
  • Capability inventory: None; the skill is explicitly 'no code' and lacks command execution or network capabilities.
  • Sanitization: Absent; the skill does not provide guidance on validating or sanitizing source material.
Audit Metadata
Risk Level
SAFE
Analyzed
Mar 7, 2026, 03:47 AM