The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill describes patterns for ingesting and storing untrusted data, such as user file uploads and form submissions, which represents a potential attack surface for indirect prompt injection.
Ingestion points: Data enters the agent's context through req.formData() and URL parameters (context.params) in Netlify Functions and Edge Functions as shown in the code examples (SKILL.md).
Boundary markers: The instructions emphasize store isolation, specifically warning that site-wide stores are shared across all deploy contexts and recommending getDeployStore() for better isolation.
Capability inventory: The skill enables reading, writing, and deleting data using the @netlify/blobs client library.
Sanitization: The documentation explicitly advises gating reads behind authenticated functions and warns against serving arbitrary user-supplied keys for sensitive data.
[EXTERNAL_DOWNLOADS]: The skill references and downloads resources from well-known technology services.
Package installation: Recommends installing the @netlify/blobs package from the public registry (SKILL.md).
Repository reference: The changelog (CHANGELOG.md) references the netlify/context-and-tools repository on GitHub as the source for the skill's workflow.
Documentation links: Provides links to the official Netlify documentation for further technical background (SKILL.md).

netlify-blobs