teach

Pass

Audited by Gen Agent Trust Hub on Jul 7, 2026

Risk Level: SAFECOMMAND_EXECUTIONPROMPT_INJECTIONNO_CODE
Full Analysis
  • [COMMAND_EXECUTION]: The skill instructions state that the agent should "open the lesson file for the user by running a CLI command." This creates a surface for potential command injection if the agent handles filenames or the execution environment unsafely.\n- [PROMPT_INJECTION]: The skill is vulnerable to indirect prompt injection because it is designed to ingest and process data from external links and user-defined mission files to generate content.\n
  • Ingestion points: The agent retrieves context from RESOURCES.md (external URLs), MISSION.md, and NOTES.md.\n
  • Boundary markers: There are no instructions providing delimiters or boundary markers to separate trusted instructions from untrusted external data.\n
  • Capability inventory: The agent has the ability to create/write files (HTML, Markdown) and execute CLI commands.\n
  • Sanitization: No validation or sanitization steps are defined for content retrieved from external sources before it is used to generate lessons or assets.\n- [NO_CODE]: The skill consists entirely of Markdown-based instructions and formatting templates; it does not include any executable scripts, binaries, or third-party code packages.
Audit Metadata
Risk Level
SAFE
Analyzed
Jul 7, 2026, 02:17 AM
Security Audit — agent-trust-hub — teach