The Agent Skills Directory

[SAFE]: The skill exclusively uses the official refly CLI tool for execution and status monitoring, which is appropriate for its stated purpose within the Refly ecosystem.
[SAFE]: No instances of obfuscation, hardcoded credentials, or malicious network exfiltration patterns were identified.
[INDIRECT_PROMPT_INJECTION]: The skill provides an interface for ingesting user-supplied video metadata.
Ingestion points: Video title, description, and tags in SKILL.md.
Boundary markers: Absent in documentation.
Capability inventory: CLI-based workflow invocation and monitoring.
Sanitization: Not explicitly defined in the skill files; platform-side handling is expected.

youtube