scaffold-exercises
Pass
Audited by Gen Agent Trust Hub on Sep 28, 2026
Risk Level: SAFECOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTIONDYNAMIC_EXECUTION
Full Analysis
- [COMMAND_EXECUTION]: The skill requires executing
mkdir,git, andpnpm ai-hero-clito scaffold files and validate their structure. - [INDIRECT_PROMPT_INJECTION]: The workflow involves parsing a plan to generate directory paths. Ingestion point: Step 1 (Parse the plan). Boundary markers: Absent. Capability inventory:
mkdir,git,pnpm. Sanitization: Absent. This ingestion of untrusted data represents an attack surface for manipulating command arguments. - [DYNAMIC_EXECUTION]: The agent is instructed to dynamically build shell commands based on the input plan to create directories and move folders.
Audit Metadata