The Agent Skills Directory

[SAFE]: The skill consists of Markdown and LaTeX documentation and reference guides. It provides structured logic for scientific writing based on established academic patterns.
[INDIRECT_PROMPT_INJECTION]: The skill processes user-provided research claims and drafts (including Chinese language notes) to generate manuscript prose.
Ingestion points: Author-provided claims, results, figures, and drafts are ingested in SKILL.md to be transformed into scientific writing.
Boundary markers: The skill recommends using explicit placeholders when evidence is missing but does not implement strict data delimiters.
Capability inventory: The skill has no functional capabilities to execute commands, access the file system, or perform network requests.
Sanitization: No explicit sanitization of user data is performed as the skill is intended for drafting text within the agent's output context only.

nature-writing