The Agent Skills Directory

[SAFE]: The skill is entirely composed of natural language instructions and templates designed to guide the user in task management and prompt engineering.
[NO_CODE]: There are no executable scripts, shell commands, or binary files associated with this skill. It operates solely within the context of text generation.
[INDIRECT_PROMPT_INJECTION]: The skill provides a framework for processing user-supplied tasks. While it ingests untrusted input from the user, it lacks any dangerous capabilities (such as file system access, network requests, or code execution tools) that could be exploited through injection. The instructions focus on outputting structured text for the user's reference.

task-drill