The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill instructions create an attack surface for indirect prompt injection by interpolating untrusted user data directly into shell command arguments.
Ingestion points: User-defined strings for [theme], [description], and [region] found in the seven priority workflows in SKILL.md.
Boundary markers: Absent. The skill provides no instructions to wrap, escape, or delimit the user input within the shell command.
Capability inventory: The skill utilizes the Bash tool to execute the ~/.claude/scripts/call-image-api script.
Sanitization: There is no evidence of input validation, shell-escaping, or filtering mechanisms to prevent malicious shell metacharacters from being executed.
[COMMAND_EXECUTION]: The skill is designed to execute a local utility script (~/.claude/scripts/call-image-api) via the Bash tool to interface with OpenAI and Google image engines.
[DYNAMIC_EXECUTION]: The skill performs dynamic assembly of shell commands at runtime by concatenating the static script path and flags with variable user-provided prompt content.

image-creator