The Agent Skills Directory

[COMMAND_EXECUTION]: The skill documents standard usage of the Docusaurus CLI (npx docusaurus start, build, swizzle, deploy). These are legitimate development operations required for the framework's functionality.
[EXTERNAL_DOWNLOADS]: The skill references standard plugins and themes from the official Docusaurus ecosystem (e.g., @docusaurus/theme-mermaid, remark-math). These are well-known, trusted packages from the Docusaurus project and common package registries.
[SAFE]: The skill provides strong security advice, such as recommending the use of search-only API keys for Algolia DocSearch and configuring onBrokenLinks: 'throw' to ensure build integrity.
[PROMPT_INJECTION]: No patterns associated with prompt injection, safety bypasses, or instruction overrides were detected in the instructions or examples.
[DATA_EXFILTRATION]: No exfiltration patterns or sensitive file access (e.g., SSH keys, credentials) were found. Placeholders are used for sensitive configuration values.
[INDIRECT_PROMPT_INJECTION]: While the skill involves processing documentation content which is a standard attack surface for indirect injection, it focuses on site generation and does not implement unsafe data ingestion or interpolation patterns that would elevate risk beyond the baseline of document processing.

web-meta-framework-docusaurus