granola-performance-tuning

Pass

Audited by Gen Agent Trust Hub on Mar 25, 2026

Risk Level: SAFE
Full Analysis
  • [SAFE]: The skill provides purely instructional content focused on audio setup, meeting practices, and template optimization for transcription services.
  • [SAFE]: External links in the documentation point to the official domain of the Granola service (granola.ai), which is appropriate for the skill's purpose.
  • [SAFE]: No executable code, remote script downloads, obfuscation, or sensitive data access patterns were found in the analyzed files.
  • [PROMPT_INJECTION]: The skill defines processes for handling external data (meeting transcriptions), which constitutes a surface for indirect prompt injection. However, the guidelines focus on improving structural clarity to aid parsing rather than introducing exploitable logic.
  • Ingestion points: Meeting transcriptions and note data referenced in SKILL.md and references/implementation.md.
  • Boundary markers: The skill encourages the use of clear Markdown headers and structured templates to demarcate sections.
  • Capability inventory: The skill requests standard file interaction permissions (Read, Write, Edit).
  • Sanitization: Not applicable as the skill provides high-level workflow guidance.
Audit Metadata
Risk Level
SAFE
Analyzed
Mar 25, 2026, 04:41 PM