The Agent Skills Directory

[REMOTE_CODE_EXECUTION]: The skill's documentation provides a command to install the necessary CLI tool by fetching and executing a script from https://runcomfy.com/install.sh directly through a shell pipe.
[EXTERNAL_DOWNLOADS]: The skill references the @runcomfy/cli package on the NPM registry and communicates with the RunComfy service infrastructure to perform image generation and download results.
[COMMAND_EXECUTION]: The skill is configured to execute the runcomfy command, which allows the agent to interact with the host environment to run models, manage authentication, and download files. It interpolates user-supplied data into command arguments.
[PROMPT_INJECTION]: The skill handles untrusted user input that is passed to remote AI models, creating a surface for indirect prompt injection.
Ingestion points: User-provided text prompts for image generation as described in the SKILL.md file.
Boundary markers: The prompt text is encapsulated within a JSON object passed to the CLI's --input argument.
Capability inventory: The runcomfy CLI tool can perform network requests to the RunComfy API and write generated image files to the local file system.
Sanitization: No explicit content filtering or validation of the user's prompt is implemented beyond the JSON encapsulation.

ai-image-generation