The Agent Skills Directory

[COMMAND_EXECUTION]: The skill uses npx hyperframes for lifecycle tasks including skill updates, authentication status checks, project initialization, linting, and final project validation checks.
[INDIRECT_PROMPT_INJECTION]: The skill ingests untrusted data from external Figma URLs and local project files such as BRIEF.md, STORYBOARD.md, and hyperframes.json. This represents an attack surface where malicious content could attempt to influence the agent's behavior. The skill documents specific mitigation strategies, such as avoiding raw Figma connector calls in favor of adapters that perform SVG sanitization.
[DYNAMIC_EXECUTION]: The scripts/frame-packets.mjs script dynamically assembles instruction sets for sub-agents by concatenating core framework markdown files with skill-specific logic. These assembled instructions are then used to drive worker contexts during the build phase.
[REMOTE_CODE_EXECUTION]: The skill uses npx to execute the hyperframes package from the npm registry. While this is a standard practice for this toolchain, it involves downloading and executing remote code at runtime.

general-video