The Agent Skills Directory

[COMMAND_EXECUTION]: The skill executes the omc ask command by interpolating dynamic arguments ({{ARGUMENTS}}). This pattern introduces a surface for shell command injection if the arguments are derived from untrusted sources and are not properly escaped before execution.
[PROMPT_INJECTION]: The skill routes data to local LLM providers via CLI tools, creating an indirect prompt injection surface. 1. Ingestion points: The {{ARGUMENTS}} placeholder in SKILL.md is the primary entry point for untrusted data. 2. Boundary markers: There are no boundary markers or instructions to ignore embedded commands in the template. 3. Capability inventory: The skill can execute subprocesses (omc, claude, codex, gemini) and write files to .omc/artifacts/. 4. Sanitization: The skill does not implement any sanitization or validation of the input arguments.

ask