The Agent Skills Directory

[PROMPT_INJECTION]: The skill instructions define operational logic for adding music based on video context. No instructions were found that attempt to bypass safety guidelines, override system behavior, or extract sensitive prompts.
[DATA_EXPOSURE_EXFILTRATION]: No evidence of sensitive file access or unauthorized network operations. The skill uses a proprietary 'music' operation to interact with a stock media library.
[REMOTE_CODE_EXECUTION]: The skill mentions the use of the vendor-provided @scrollmark/cli for running the editor in an MCP context. This is a legitimate vendor resource and does not involve arbitrary remote code execution from untrusted sources.
[DYNAMIC_EXECUTION]: The skill does not generate or execute code at runtime. It constructs arguments for a predefined 'music' tool based on user-provided vibes.
[INDIRECT_PROMPT_INJECTION]: The skill ingests user input to formulate search queries for music. While this is a data ingestion point, the capability is restricted to searching a stock library and does not present an exploitable attack surface for privilege escalation or data theft.

editor-soundtrack