The Agent Skills Directory

[DYNAMIC_EXECUTION]: The skill uses a custom bootstrapping function (importPackagesOrBootstrap) in animation-map.mjs and contrast-report.mjs to dynamically install and load NPM packages such as @hyperframes/producer and sharp at runtime.
[COMMAND_EXECUTION]: The Python script extract-audio-data.py executes the ffmpeg system binary via subprocess.run to extract audio information from media files. The skill also makes heavy use of the npx hyperframes CLI for rendering and validation.
[EXTERNAL_DOWNLOADS]:
The skill references an external repository (pixel-point/animate-text) for text animation effects, instructing the agent to install it via npx skills add.
Multiple animation libraries (GSAP, Lottie, Three.js, Anime.js) are loaded from public CDNs like cdn.jsdelivr.net, cdnjs.cloudflare.com, and unpkg.com within HTML templates.
[INDIRECT_PROMPT_INJECTION]: The skill exposes a broad attack surface for indirect prompt injection through its capability to ingest untrusted content from external sources.
Ingestion points: Content is ingested from external URLs using the hyperframes capture command as described in engine/hyperframes-cli/GUIDE.md.
Boundary markers: None detected; no clear delimiters or instructions to ignore embedded commands are present in the provided documentation for processing content from captured sites.
Capability inventory: The skill possesses capabilities to execute shell commands, read and write files, and perform network operations.
Sanitization: There are no instructions for sanitizing or filtering the content retrieved from external sources before it enters the agent's context.

recoup-internal-consulting-hyperframes-video