The Agent Skills Directory

[SAFE]: The skill consists of instructional markdown guidelines for AI behavior. It does not include executable code or shell commands that run autonomously in the agent's environment.- [SAFE]: The YAML frontmatter is used correctly to scope the skill without requesting unnecessary permissions or access to sensitive tools.- [SAFE]: The repository provides a high degree of transparency by including detailed research notes, blind validation results, and adversarial review documentation, all of which prioritize factual accuracy and safe usage.- [SAFE]: External dependencies and tools mentioned in the repository (such as Next.js, Cloudflare Wrangler, and the 'measure.py' script) are used for building documentation and metrics, and do not present a runtime security risk to the agent.- [SAFE]: URL references are restricted to reputable academic platforms (arXiv, PNAS), news organizations (The Economist, The Atlantic), and established open-source projects on GitHub.

humanizing-writing