The Agent Skills Directory

[SAFE]: The skill defines a structured workflow for generating and editing images using native tools (such as GPT Image 2). It includes detailed rules for capability detection and output handling.
[COMMAND_EXECUTION]: The skill includes a utility script, scripts/validate-image-maker.mjs, designed to run in a Bun environment. This script performs file system operations (read, write, delete) but implements robust security measures including path normalization, symlink detection, and the use of the O_NOFOLLOW flag to prevent directory traversal attacks. These operations are restricted to temporary directories or the specific .hyper/image-maker/ path.
[DATA_EXFILTRATION]: No network operations or patterns associated with unauthorized data transmission were found. Sensitive file paths are not accessed, and file writing is strictly scoped to designated subdirectories.
[PROMPT_INJECTION]: The instructions are focused on task execution, output formatting, and error handling. There are no attempts to override system safety guidelines or extract system prompts.
[SAFE]: The skill includes security-conscious validation logic, such as the validateTopic function, which filters control characters and ensures directory names are safe before use.

image-maker