create-verification-skill
Pass
Audited by Gen Agent Trust Hub on Oct 2, 2026
Risk Level: SAFECOMMAND_EXECUTIONDYNAMIC_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill instructs the agent to discover and run build or start commands found in the target repository's configuration (e.g., package scripts, Makefiles).
- [DYNAMIC_EXECUTION]: The skill generates a new agent skill and mandates an immediate execution of its instructions to verify the generation was successful.
- [INDIRECT_PROMPT_INJECTION]: The skill ingests untrusted codebase content to drive its generation logic, creating a surface for repository-based prompt injection.
- Ingestion points: Files within the target repository, including READMEs, routes, and configuration files.
- Boundary markers: No explicit delimiters are used to distinguish repository data from instructions.
- Capability inventory: Discovery and execution of shell commands and file system writes to the skill directory.
- Sanitization: No sanitization is performed on codebase strings before they are interpolated into the generated skill.
Audit Metadata