The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill ingests untrusted data from external video sources and transcripts, creating a vulnerability surface for the agent.
Ingestion points: External data enters the context through opusclip project create and transcript retrieval commands as documented in SKILL.md.
Boundary markers: No specific boundary markers or instructions are defined to isolate untrusted data from the agent's instructions.
Capability inventory: The skill executes a bundled CLI utility, invokes ffmpeg, and performs local file writes to generate previews.
Sanitization: There is no mention of sanitizing transcript data before it is presented to the agent for processing.
[COMMAND_EXECUTION]: The skill utilizes a bundled script scripts/opusclip and requires the external utility ffmpeg to perform core media tasks.
[DYNAMIC_EXECUTION]: The skill generates an HTML preview file using templates/preview.html by interpolating external project data, which constitutes local script generation from templates.

opusclip