perf-pass
Pass
Audited by Gen Agent Trust Hub on Oct 29, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTIONCOMMAND_EXECUTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill ingests untrusted data from the project's own files and live browser sessions which could theoretically contain malicious instructions.
- Ingestion points: Reads
src/budgets.ts,3D-STYLE.md, and the output of theledger()profiling tool from a running browser session. - Boundary markers: No explicit delimiters or instructions are specified to distinguish between project data and agent instructions.
- Capability inventory: The skill can modify project source files (
src/budgets.ts,3D-STYLE.md,AGENTS.md), run development servers (vite preview), execute testing frameworks (vitest), and run CLI tools (three-studio). - Sanitization: There is no mention of sanitizing or validating the content of the budget files or profiling output before it influences the agent's actions.
- [COMMAND_EXECUTION]: The skill automates the execution of various CLI tools and development commands to verify fixes, including
vite,lint, and project-specific testing commands.
Audit Metadata