The Agent Skills Directory

[SAFE]: No security issues detected. The skill follows best practices for task orchestration, including model tiering and differentiating between foreground and background task execution.
[PROMPT_INJECTION]: The skill functions as a coordinator that passes user-defined task descriptions to subagents. This creates a standard surface for indirect prompt injection common to agentic workflows, though no specific malicious patterns are present.
Ingestion points: User-provided task descriptions are passed into the prompt parameter of the Task tool across multiple examples in SKILL.md.
Boundary markers: None are explicitly defined to separate user input from the subagent instructions.
Capability inventory: The skill utilizes the oh-my-claudecode:executor subagent which has the capability to perform file system operations and execute shell commands (e.g., npm install).
Sanitization: The skill does not implement explicit sanitization or input validation for the delegated task prompts, relying instead on lightweight verification steps (build/test checks) after execution.

ultrawork