scaffold-exercises
Pass
Audited by Gen Agent Trust Hub on Jul 11, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: The skill follows legitimate developer workflows for project scaffolding and directory management.
- [COMMAND_EXECUTION]: The skill instructs the agent to use local commands such as
mkdir,git mv, andpnpm. These operations are standard for the described task of setting up directory structures and are performed within the expected project scope. - [INDIRECT_PROMPT_INJECTION]: The skill ingests a user-provided 'plan' to determine directory and file names. While this is a data ingestion surface, the resulting actions (creating directories and readmes) are low-risk. Mitigation involves the agent's internal validation of file paths.
- [EXTERNAL_DOWNLOADS]: The skill references
pnpm ai-hero-cli, which appears to be a tool local to the project environment rather than a package downloaded from an untrusted remote source.
Audit Metadata