The Agent Skills Directory

[EXTERNAL_DOWNLOADS]: The skill downloads and updates its core CLI tool and workflow-specific plugins from external registries like npm and GitHub.
SKILL.md uses npx hyperframes@latest upgrade and npx hyperframes skills update to maintain tool freshness.
references/skill-lifecycle.md describes adding skills from the heygen-com/hyperframes repository.
[COMMAND_EXECUTION]: Extensive use of shell commands to drive the video production pipeline and manage dependencies.
scripts/plugin-cli.mjs acts as a launcher for the hyperframes CLI tool via npx.
Uses the gh CLI in references/routes/pr-to-video.md to retrieve pull request metadata for video generation.
[INDIRECT_PROMPT_INJECTION]: The skill ingests untrusted data from multiple external sources which could potentially contain malicious instructions meant to influence the agent.
Ingestion points: The skill accepts website URLs, GitHub PRs, and Figma designs as inputs in SKILL.md and various route files.
Boundary markers: The framework enforces a strict structure using BRIEF.md and STORYBOARD.md to isolate user intent from the technical execution of sub-agents.
Capability inventory: The skill can execute shell commands (npx), write files (HTML compositions), and perform network operations (website capture).
Sanitization: references/intent-interview.md explicitly mentions using specialized tools for Figma imports to ensure SVG sanitization and brand-token binding.
[DYNAMIC_EXECUTION]: The skill generates executable HTML and JavaScript content at runtime for video rendering.
references/frame-worker-core.md defines the contract for sub-agents to author HTML <template> fragments containing GSAP animation logic registered to the global window object.

hyperframes