The Agent Skills Directory

[SAFE]: The skill serves as a guide for interacting with Vercel Sandbox. It provides structured templates for sub-agent prompts to research official Vercel documentation.
[EXTERNAL_DOWNLOADS]: The skill instructs the agent to fetch documentation from official Vercel domains (e.g., https://vercel.com/docs/vercel-sandbox.md). These references target a well-known service and are documented neutrally for research purposes.
[INDIRECT_PROMPT_INJECTION]: The skill defines a research process that ingests data from external documentation URLs. This is the primary purpose of the skill, and the sources are restricted to a trusted service provider.
Ingestion points: Documentation pages fetched via WebFetch as defined in SKILL.md prompt templates.
Boundary markers: Uses Markdown headers to structure sub-agent context.
Capability inventory: Research tasks using WebFetch and interaction with the Vercel Sandbox SDK.
Sanitization: Standard agent processing of documentation content.

vercel-sandbox