The Agent Skills Directory

[EXTERNAL_DOWNLOADS]: The skill references an installation guide hosted on the inference-sh GitHub repository (https://raw.githubusercontent.com/inference-sh/skills/refs/heads/main/cli-install.md). This is documented neutrally as a standard setup procedure for the vendor's tool.
[COMMAND_EXECUTION]: Instructs the user to run belt login and belt app run commands to interact with the upscaling service. These commands are part of the intended functionality and are restricted to the belt CLI tool via the skill's tool definitions.
[INDIRECT_PROMPT_INJECTION]: The skill is designed to process user-provided image URLs which are passed to the belt CLI.
Ingestion points: The image_url field within the JSON input for the belt app run command (SKILL.md).
Boundary markers: No explicit boundary markers or safety instructions for the agent regarding external input were found in the skill text.
Capability inventory: The skill uses the Bash(belt *) tool to execute commands, including those that process the external URLs (SKILL.md).
Sanitization: Input validation is not performed within the skill instructions, relying on the underlying belt tool and inference.sh backend.

image-upscaling