The Agent Skills Directory

[COMMAND_EXECUTION]: The skill makes extensive use of shell commands to orchestrate the video pipeline, including python3, ffprobe, npm, and npx for tasks ranging from media probing to rendering and QA.
[EXTERNAL_DOWNLOADS]: The skill utilizes npx to execute the hyperframes engine and includes instructions to perform remote updates via npx hyperframes skills update, which involves fetching and executing content from external registries.
[INDIRECT_PROMPT_INJECTION]: The skill processes data from external sources, which constitutes an attack surface for indirect prompt injection:
Ingestion points: Fetches reviewer notes from external URLs via video-review-canvas and processes user-supplied reference reels.
Boundary markers: There are no explicit delimiters or instructions to ignore embedded commands in the ingested data.
Capability inventory: The agent can execute shell commands (python3, npm, npx) and perform network operations, which could be misused if malicious instructions are injected via the review notes or style references.
Sanitization: No sanitization or validation of external content (notes/references) is specified in the instructions.

video-edit-pipeline