sanity-visual-coverage
Pass
Audited by Gen Agent Trust Hub on Oct 8, 2026
Risk Level: SAFECOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill instructs the agent to run project-specific commands like
pnpm visual-coverageand GitHub CLI commands such asgh pr list. These are standard developer operations required for the skill's operation and are targeted at the project's own repository. - [INDIRECT_PROMPT_INJECTION]: The skill analyzes data from external sources (GitHub PR titles and file lists), which represents a potential injection surface.
- Ingestion points: External data is ingested from GitHub using
gh pr listandgh pr checkoutto identify pending stories and PR statuses. - Boundary markers: The instructions do not define specific markers to delimit untrusted content from the PR metadata.
- Capability inventory: The skill utilizes
pnpmfor local script execution andghfor interacting with the GitHub API. - Sanitization: No specific sanitization or filtering of PR metadata is described in the prompt logic, relying on standard CLI tool output parsing.
Audit Metadata