Skills
Navigate
skills/netlify/context-and-tools/netlify-image-cdn/Gen Agent Trust Hub

netlify-image-cdn

Pass

Audited by Gen Agent Trust Hub on Sep 28, 2026

Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
  • [INDIRECT_PROMPT_INJECTION]: The skill describes a user-upload pipeline in references/user-uploads.md that processes external data.
  • Ingestion points: The netlify/functions/upload.ts script ingest untrusted data from client-side multipart/form-data requests via req.formData().
  • Boundary markers: As the ingestion involves binary image data, no specific prompt delimiters or boundary markers are employed.
  • Capability inventory: The skill demonstrates writing to persistent storage using @netlify/blobs and serving/transforming that data via Netlify Image CDN through [[redirects]] in netlify.toml.
  • Sanitization: The reference code implements standard validation by checking the file's MIME type against an ALLOWED_TYPES list and enforcing a MAX_SIZE limit of 4MB.
Audit Metadata
Risk Level
SAFE
Analyzed
Sep 28, 2026, 04:24 PM