design-goal
Pass
Audited by Gen Agent Trust Hub on Oct 14, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTIONCOMMAND_EXECUTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill reads and processes external data from the project repository, including source code, component definitions, and styling files. This establishes an ingestion surface where instructions embedded in project files could potentially influence the agent's behavior. * Ingestion points: Project styling systems, components, route templates, and layout files (SKILL.md, Section 1). * Boundary markers: None present to distinguish between instructions and data. * Capability inventory: File write access for UI refinements and process execution for builds, tests, linting, and browser tooling (SKILL.md, Sections 4 and 7). * Sanitization: No explicit sanitization or filtering of project content.
- [COMMAND_EXECUTION]: The skill involves executing local development tools, including application servers, screenshot utilities, and build pipelines (type checks, linters, and tests). These operations are standard for a development-focused agent and are scoped to the project environment.
Audit Metadata