The Agent Skills Directory

[SAFE]: The skill is designed to improve research integrity by using structured templates and verification protocols to prevent AI-generated hallucinations and unverified claims.
[COMMAND_EXECUTION]: The skill includes a local script scripts/draft_lint.py for automated checking of paper drafts and recommends using standard document conversion tools like markitdown. These are intended for local research processing.
[EXTERNAL_DOWNLOADS]: The documentation references legitimate academic tools and repositories, such as scientific-writer and markitdown from well-known sources, as part of its recommended research workflow.
[DATA_EXFILTRATION]: The citation verification process uses curl to interact with standard academic APIs like Crossref and arXiv to verify metadata. These actions are for research integrity and do not target sensitive user data.

light-paper-drafting