The Agent Skills Directory

[SAFE]: The skill actively promotes security best practices, including detailed guidance on Row-Level Security (RLS), authentication, and the principle of least privilege for database functions.
[SAFE]: All external references and network operations target official Supabase domains and repositories, such as supabase.com and the supabase/agent-skills GitHub repository, which are recognized as well-known and trusted sources.
[INDIRECT_PROMPT_INJECTION]: The skill ingests data from official documentation and changelogs to influence its behavior. This surface is considered safe as it relies exclusively on official vendor sources.
Ingestion points: Fetches content from https://supabase.com/changelog.md and documentation pages via supabase.com/docs/*.md (SKILL.md).
Boundary markers: No explicit delimiters are specified for the fetched data.
Capability inventory: The skill uses database tools like execute_sql and the supabase CLI, and possesses the capability to create GitHub issues for user feedback (references/skill-feedback.md).
Sanitization: No explicit sanitization of fetched documentation content is mentioned, as the content originates from the official vendor.

supabase