The Agent Skills Directory

[SAFE]: No malicious patterns or security vulnerabilities were identified in the skill. All operations are aligned with the stated goal of documentation automation and follow established development workflows.\n- [COMMAND_EXECUTION]: The skill uses standard npm commands, such as npm ci, npm run build, and npm run docs:dev, which are required for managing and building Docusaurus sites. These commands are executed within the project directory to perform legitimate build and development tasks.\n- [EXTERNAL_DOWNLOADS]: References official and trusted sources, including the Vercel configuration schema (openapi.vercel.sh) and the NPM registry for the @docusaurus/theme-mermaid package. No untrusted or unknown third-party repositories are utilized.\n- [PROMPT_INJECTION]: Analysis of the instructions and bundled references found no evidence of malicious prompt injection, such as attempts to bypass safety filters or override agent constraints.\n- [DATA_EXFILTRATION]: While the skill inventories project files to generate documentation, this behavior is restricted to codebase documentation purposes and does not involve accessing sensitive system credentials or exfiltrating data to external domains.

docs-platform-rollout