The Agent Skills Directory

[COMMAND_EXECUTION]: The skill instructs the agent to use standard tools like ls, cat, rg, and grep to inspect the local filesystem for project configuration and architecture signatures (e.g., checking package.json, schema.prisma, and supabase/config.toml). These operations are consistent with the skill's stated purpose of auditing backend architecture.
[EXTERNAL_DOWNLOADS]: Code examples for Supabase Edge Functions reference external module imports from esm.sh (https://esm.sh/@supabase/supabase-js@2). esm.sh is a well-known and widely used CDN for JavaScript modules, and this represents standard development practice.
[INDIRECT_PROMPT_INJECTION]: The skill ingests structural data from project files such as database schemas and package manifests. While these are external inputs, they are used for architectural classification and code generation rather than being interpolated into prompts in a manner that would override agent safety guidelines.

backend-patterns