scaffold-exercises

Pass

Audited by Gen Agent Trust Hub on Sep 20, 2026

Risk Level: SAFECOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
  • [COMMAND_EXECUTION]: The skill instructs the agent to use shell commands including mkdir -p, git commit, git mv, and pnpm ai-hero-cli to scaffold and validate directory structures. These operations are consistent with the tool's purpose as a developer utility.\n- [INDIRECT_PROMPT_INJECTION]: The skill possesses a surface for indirect prompt injection by processing untrusted course plans.\n
  • Ingestion points: User-provided plans described in the 'Workflow' section (SKILL.md).\n
  • Boundary markers: None specified in the instructions.\n
  • Capability inventory: File system access (mkdir), file writing (readme.md), and command execution (pnpm, git).\n
  • Sanitization: None specified; relies on the agent's parsing logic.
Audit Metadata
Risk Level
SAFE
Analyzed
Sep 20, 2026, 07:48 AM
Security Audit — agent-trust-hub — scaffold-exercises