create-verification-skill
Pass
Audited by Gen Agent Trust Hub on Sep 15, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTIONCOMMAND_EXECUTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill is susceptible to indirect prompt injection because it instructs the agent to ingest and analyze untrusted third-party repository data (documentation, configuration files, and package scripts) to determine application behavior and generate verification instructions.
- Ingestion points: Repository source code, Makefiles, package scripts, and README files.
- Boundary markers: Absent; there are no instructions or delimiters defining boundaries to separate untrusted codebase text from structural agent instructions.
- Capability inventory: The skill performs file system modifications (creating
<skill-root>/verify-<app>/SKILL.mdand feature maps) and executes arbitrary shell commands to launch, test, and clean up the application. - Sanitization: Absent; the skill does not sanitize or validate commands extracted from the codebase before executing them.
- [COMMAND_EXECUTION]: The skill requires the agent to run commands discovered within the analyzed repository, such as the codebase's documented dev or startup commands. Running unverified commands from an untrusted repository can result in arbitrary command execution on the host machine during the testing and verification phase.
Audit Metadata