The Agent Skills Directory

[SAFE]: No security issues detected. The skill provides instructions for interacting with the FilmCraft application through its official, documented automation interfaces.
[INDIRECT_PROMPT_INJECTION]: The skill processes project data and UI state, creating a standard indirect prompt injection surface. * Ingestion points: Data enters the agent context through project_inspect, ui.inspect, and sequence_inspect (SKILL.md). * Boundary markers: No specific delimiters or safety warnings for untrusted project metadata are defined in the instructions. * Capability inventory: The agent can modify timelines via engine.execute and write files via file.exportMedia (control-protocol.md). * Sanitization: The skill relies on the application's structured JSON protocol for data handling; no additional prompt-level sanitization is specified.

filmcraft