netlify-blobs
Pass
Audited by Gen Agent Trust Hub on Sep 28, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill is designed to handle untrusted data such as file uploads and JSON payloads which could theoretically contain malicious instructions targeting the processing agent.
- Ingestion points: Untrusted data enters the context via
req.formData()in the 'Persist an upload' example and throughstore.get()operations inSKILL.md. - Boundary markers: The skill provides explicit 'Footguns' documentation advising users to treat input as unsafe and to scope keys to prevent tampering.
- Capability inventory: The skill enables file write, read, and delete operations using the
@netlify/blobslibrary. - Sanitization: The documentation recommends gating reads behind authenticated functions and avoiding the use of arbitrary caller-supplied keys.
Audit Metadata