scaffold-exercises

Pass

Audited by Gen Agent Trust Hub on Sep 15, 2026

Risk Level: SAFEINDIRECT_PROMPT_INJECTIONCOMMAND_EXECUTION
Full Analysis
  • [INDIRECT_PROMPT_INJECTION]: The skill processes a user-provided plan to generate directory names and file contents, which introduces a surface for indirect prompt injection. \n
  • Ingestion points: SKILL.md workflow step "Parse the plan".\n
  • Boundary markers: No specific delimiters or warnings for the agent to ignore instructions embedded in the plan are defined.\n
  • Capability inventory: mkdir, file creation (readme.md), git commit, git mv, and pnpm ai-hero-cli internal lint in SKILL.md.\n
  • Sanitization: The skill enforces specific naming patterns (dash-case and numeric prefixes) which limits the scope of directory creation.\n- [COMMAND_EXECUTION]: The skill uses shell commands for directory creation, version control, and project linting.
Audit Metadata
Risk Level
SAFE
Analyzed
Sep 15, 2026, 03:01 AM
Security Audit — agent-trust-hub — scaffold-exercises