optimize
Pass
Audited by Gen Agent Trust Hub on Sep 15, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill ingests user-provided descriptions of performance issues and component names to provide optimization advice. This creates an attack surface where malicious instructions embedded in performance reports could potentially influence agent behavior.
- Ingestion points: The
[route, component, or performance issue]argument in the skill definition (SKILL.md). - Boundary markers: The instructions do not define specific delimiters to isolate user input or warnings to ignore embedded instructions.
- Capability inventory: The skill instructs the agent to analyze and modify project source code (HTML, CSS, JavaScript) based on optimization guidelines.
- Sanitization: No input validation or sanitization routines are specified for the user-provided data.
- [SAFE]: The skill promotes standard, high-quality performance optimization practices such as image format selection (AVIF/WebP), font subsetting, and minimizing layout thrashing. It references reputable industry tools and avoids dangerous command execution or remote script downloads.
Audit Metadata