create-verification-skill
Pass
Audited by Gen Agent Trust Hub on Oct 9, 2026
Risk Level: SAFECOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTIONDYNAMIC_EXECUTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill analyzes external data from the repository (codebase, READMEs, and build files) to generate verification instructions. There are no explicit sanitization steps or boundary markers defined to prevent the agent from adopting malicious instructions embedded in the repository's documentation.
- Ingestion points: Analyzes repository files including README, Makefile, and package scripts in SKILL.md.
- Capability inventory: Executes local commands identified in the repo and writes new skill files to the .cursor/skills/ directory as described in SKILL.md.
- Boundary markers: None identified.
- Sanitization: None identified.
- [COMMAND_EXECUTION]: The skill is instructed to discover and run existing commands from the repository (e.g., make, npm scripts, flash tools). While this is intended for verification purposes, it enables the execution of arbitrary commands defined within the codebase being analyzed.
- [DYNAMIC_EXECUTION]: The skill generates a project-local agent skill (.cursor/skills/verify-/SKILL.md) at runtime. This generated content becomes a new set of executable instructions for the agent, based on findings from the initial repository analysis.
Audit Metadata