charles-learn-anything

Pass

Audited by Gen Agent Trust Hub on Nov 19, 2026

Risk Level: SAFECOMMAND_EXECUTIONEXTERNAL_DOWNLOADSCREDENTIALS_UNSAFEINDIRECT_PROMPT_INJECTION
Full Analysis
  • [COMMAND_EXECUTION]: The kit/build.py script executes shell commands to process audio files using ffmpeg and to open the generated lesson in Google Chrome. These operations are essential for the skill's build-and-view workflow.
  • [EXTERNAL_DOWNLOADS]: The build script performs network requests to the fal.ai API to generate and download audio narration files. This functionality is only activated if the user provides an API key and is a core feature for high-quality text-to-speech.
  • [CREDENTIALS_UNSAFE]: The skill implements a mechanism to locate the FAL_KEY API credential by searching for .env files in the current directory and its parent directories. This is a standard convenience feature for local development tools and does not involve exfiltration.
  • [INDIRECT_PROMPT_INJECTION]: The skill processes user-supplied topics to generate interactive content rendered in a browser. While this presents an attack surface for indirect injection, the risk is mitigated by the agent's strict instructions to generate the content itself following a specific educational framework.
Audit Metadata
Risk Level
SAFE
Analyzed
Nov 19, 2026, 12:40 PM