design-normalize
Pass
Audited by Gen Agent Trust Hub on Nov 2, 2026
Risk Level: SAFECOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill instructs the agent to run project build, linter, and tests in Phase 9 (Self-Healing Validation). This is a standard and expected operation for a development agent ensuring code integrity after automated refactoring.
- [INDIRECT_PROMPT_INJECTION]: The skill scans the entire codebase to identify styling deviations, which introduces a potential surface for indirect prompt injection if malicious instructions are hidden within code comments or metadata.
- Ingestion points: codebase-wide scan of CSS, SCSS, JS, TS, Dart, Swift, and Kotlin files as defined in Phases 1 and 2.
- Boundary markers: The skill does not explicitly define delimiters for processed code, but its operations are constrained to identifying specific design-related patterns such as hex codes and pixel values.
- Capability inventory: Reading and writing to the local filesystem and executing build/test commands as documented in the normalization and validation phases.
- Sanitization: No specific input sanitization is mentioned, relying on the pattern-matching logic to filter relevant design data.
Audit Metadata