teach

Pass

Audited by Gen Agent Trust Hub on Aug 30, 2026

Risk Level: SAFEINDIRECT_PROMPT_INJECTIONDYNAMIC_EXECUTIONCOMMAND_EXECUTION
Full Analysis
  • [INDIRECT_PROMPT_INJECTION]: The skill logic involves reading several user-modifiable files (MISSION.md, RESOURCES.md, NOTES.md, and learning records) to determine teaching content, creating a risk that malicious data in these files could influence agent behavior.
  • Ingestion points: MISSION.md, RESOURCES.md, NOTES.md, and ./learning-records/*.md.
  • Boundary markers: Absent. There are no instructions to the agent to treat data in these files as untrusted or to use specific delimiters.
  • Capability inventory: The agent has the ability to write files to the workspace and execute CLI commands to open files.
  • Sanitization: Absent. The skill does not provide instructions for sanitizing or escaping content retrieved from external resources or user files.
  • [DYNAMIC_EXECUTION]: The skill generates and manages HTML files containing interactive components like simulators and quiz widgets (JavaScript) in the ./assets/ and ./lessons/ directories. This involves the agent generating and storing executable code on the local filesystem.
  • [COMMAND_EXECUTION]: The instructions suggest the agent run a CLI command to open generated lesson files for the user. This involves the agent executing shell commands based on its own generated file paths.
Audit Metadata
Risk Level
SAFE
Analyzed
Aug 30, 2026, 06:39 PM