The Agent Skills Directory

[COMMAND_EXECUTION]: The skill instructs the agent to execute various CLI commands using npx, including hyperframes skills update, hyperframes catalog, hyperframes add, hyperframes present, hyperframes lint, and hyperframes check. These tools are part of the core HyperFrames ecosystem and are used for updating skills, project management, and runtime validation.
[EXTERNAL_DOWNLOADS]: The skill fetches resources and code from external sources: (1) Downloads the Three.js library from the JSDelivr CDN, a well-known and trusted service for front-end assets. (2) Utilizes npx to download and run the hyperframes package from the official npm registry.
[INDIRECT_PROMPT_INJECTION]: The skill processes untrusted external data, creating an attack surface for indirect prompt injection. (1) Ingestion points: The agent is instructed to process content from Figma URLs and to port existing source web pages (including DOM, CSS, and JS) into slideshow frames (SKILL.md). (2) Boundary markers: There are no specific instructions or delimiters defined to isolate the agent from potential malicious prompts embedded within the Figma source or the targeted web page's content. (3) Capability inventory: The agent has the ability to execute CLI commands via npx, generate and modify code, and access the local file system within the project directory. (4) Sanitization: Documentation mentions that bypassing specific tools for Figma imports skips SVG sanitization, implying that while sanitization features exist, they rely on the agent following the correct tool path.

hyperframes-slideshow