The Agent Skills Directory

[COMMAND_EXECUTION]: The skill utilizes the belt CLI (from inference.sh) to execute various AI-powered applications. These include falai/flux-dev-lora for image generation, x/post-create for publishing to X (Twitter), google/veo-3-1-fast for video creation, and falai/dia-tts for text-to-speech conversion. All commands are scoped within the Bash(belt *) tool restriction.
[EXTERNAL_DOWNLOADS]: The documentation provides links to installation instructions for the required CLI tool hosted on GitHub (inference-sh/skills). This is a prerequisite for the skill's functionality.
[PROMPT_INJECTION]: The skill is susceptible to indirect prompt injection (Category 8). It processes external data sources such as blog posts and podcast transcripts and interpolates them into prompts for AI generation and social media posting without explicit sanitization or boundary markers.
Ingestion points: Data is ingested from long-form source content like blog posts, podcast transcripts, and video scripts (referenced in SKILL.md).
Boundary markers: There are no specific boundary markers or instructions to ignore embedded commands in the templates provided for repurposing.
Capability inventory: The skill has the capability to write to external platforms (x/post-create) and generate various media assets via AI apps (belt app run).
Sanitization: The skill lacks sanitization or validation steps for the input content before it is passed to the AI models or the posting tool.

content-repurposing