start-core/deployment

Pass

Audited by Gen Agent Trust Hub on Jun 6, 2026

Risk Level: SAFEEXTERNAL_DOWNLOADSCOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
  • [EXTERNAL_DOWNLOADS]: The skill recommends installing dependencies from well-known services and organizations, such as Cloudflare (@cloudflare/vite-plugin, wrangler) and Netlify (@netlify/vite-plugin-tanstack-start). It also mentions installing the nightly build of Nitro (nitro@npm:nitro-nightly@latest), which is a common practice for accessing early framework features.
  • [COMMAND_EXECUTION]: Provides standard deployment and build commands, including 'npx wrangler login', 'npx netlify deploy', and 'npm run build'. These commands are expected for the administrative task of deploying a web application.
  • [PROMPT_INJECTION]: The SEO and Head Management sections identify an indirect prompt injection surface where data from a loader function is interpolated into HTML metadata tags and JSON-LD scripts.
  • Ingestion points: External data enters the context via the 'loader' function and is utilized by the 'head' property in SKILL.md.
  • Boundary markers: None are explicitly used for the basic meta tag interpolation examples.
  • Capability inventory: The skill facilitates the creation of HTML head elements and structured data scripts.
  • Sanitization: The skill demonstrates the use of JSON.stringify for JSON-LD scripts, which helps ensure the resulting data is correctly formatted and reduces the risk of script-based injection.
Audit Metadata
Risk Level
SAFE
Analyzed
Jun 6, 2026, 03:41 AM
Security Audit — agent-trust-hub — start-core/deployment