The Agent Skills Directory

[SAFE]: The skill is composed entirely of instructional text and linguistic reference material. It establishes a complex set of rhetorical rules (e.g., adversarial reframes, trilateral root notation, build-detonate rhythm) for text generation within the agent's internal reasoning process.
[NO_CODE]: The skill does not include any executable scripts, binary files, or external package dependencies. Its operation is limited to modifying the agent's output register through provided markdown-based guidelines.
[SAFE]: The skill processes user-supplied topics through a specific command structure, but lacks any tools or system access capabilities that would allow for data exfiltration or privilege escalation.

tamarru