create-formal-verification-skill
Pass
Audited by Gen Agent Trust Hub on Nov 11, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTIONDYNAMIC_EXECUTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill instructs the agent to analyze the project's codebase to select verification targets. This processes untrusted data from the repository that could contain malicious instructions intended to manipulate the agent during the generation of the verification skill.
- Ingestion points: The agent reads files and functions from the repository to identify targets for verification (as described in the "Choose the target from the code" section of SKILL.md).
- Boundary markers: The instructions do not define boundary markers or explicit safety prompts to ignore instructions embedded within the codebase being analyzed.
- Capability inventory: The skill can write new markdown and script files to the
.agents/skills/directory and create symbolic links in the.claude/skills/directory. - Sanitization: There are no instructions for sanitizing or validating the content extracted from the repository before it is used to generate new skill content.
- [DYNAMIC_EXECUTION]: The skill's primary function is to generate new executable skill instructions (
.agents/skills/<name>/SKILL.md) at runtime and instructs the agent to immediately test and follow these generated instructions ("Follow the generated skill once, as the next agent would..."). This represents a controlled form of dynamic code generation and execution.
Audit Metadata