The Agent Skills Directory

[CREDENTIALS_UNSAFE]: The skill accesses and manages sensitive authentication tokens stored in ~/.meitu/credentials.json. It also provides instructions for writing credentials to this location using the meitu config command.
[COMMAND_EXECUTION]: The skill invokes the meitu CLI tool via shell commands to perform image editing tasks (e.g., meitu image-edit).
[EXTERNAL_DOWNLOADS]: The skill requires the installation of the meitu-cli package from the npm registry to function.
[INDIRECT_PROMPT_INJECTION]: The skill accepts user-provided data (image paths or URLs) which are interpolated into the --image_list argument of a shell command.
Ingestion points: User-provided image URLs or local file paths in the Execute step.
Boundary markers: No specific delimiters or "ignore" instructions are used around the interpolated <url_or_path> variable in the shell command.
Capability inventory: The skill possesses the capability to execute the meitu CLI tool with arbitrary arguments and access the local file system for reading and writing.
Sanitization: There is no evidence of sanitization or validation of the URL/path strings before they are passed to the CLI via shell execution.

meitu-beauty