The Agent Skills Directory

[SAFE]: No malicious patterns, prompt injections, or obfuscation techniques were found in the skill's instructions or reference documents.
[COMMAND_EXECUTION]: The skill employs cueframe-cli and a set of MCP tools for composition and rendering. These are standard operations for video production and show no evidence of privilege escalation or unauthorized access.
[EXTERNAL_DOWNLOADS]: The documentation points to https://docs.cueframe.ai for setup instructions, which is the official vendor domain and considered a safe resource.
[INDIRECT_PROMPT_INJECTION]: The skill processes transcripts and composition data, creating a potential injection surface. Ingestion points: Transcripts and existing composition states are read into the context via get_media_context and get_composition. Boundary markers: The skill does not explicitly define delimiters for untrusted transcript content. Capability inventory: The agent can perform file/state modifications via apply_composition, trigger external jobs through create_render, and capture visual data with preview_frame. Sanitization: There are no explicit instructions for sanitizing or escaping the content of processed transcripts. The risk is managed by the skill's specific focus on video production iterations.

cueframe-compose-loop