audit-performance
Pass
Audited by Gen Agent Trust Hub on Jul 24, 2026
Risk Level: SAFE
Full Analysis
- [COMMAND_EXECUTION]: The skill utilizes standard development commands like
npm run buildandnpx source-map-explorerto analyze local build artifacts and bundle sizes. These operations are performed within the scope of a performance audit. - [PROMPT_INJECTION]: The skill possesses an indirect prompt injection surface as it ingests and processes untrusted data from the web using Firecrawl search and scrape tools.
- Ingestion points: Technical research data from
firecrawl_searchandfirecrawl_scrapein SKILL.md. - Boundary markers: Absent; scraped content is used to inform the final performance audit report without explicit delimiters.
- Capability inventory: Access to build tools, file system analysis for bundle mapping, and database query explanation (
EXPLAIN ANALYZE). - Sanitization: No validation or sanitization is performed on external content fetched from the web before processing.
Audit Metadata