The Agent Skills Directory

[COMMAND_EXECUTION]: The skill facilitates the execution of various development and infrastructure commands using the Supabase CLI, GitHub CLI, Docker, and Deno to manage the deployment lifecycle.
Evidence: The skill instructs the agent to use supabase link, supabase db push, supabase functions deploy, and gh secret set in references/executing-direct-deploy.md and references/configuring-github-actions.md.
[INDIRECT_PROMPT_INJECTION]: The skill relies on reading external project files like trd.md (Product TRD) and config.toml to extract configuration parameters, creating a potential surface for indirect instruction injection.
Ingestion points: The agent is instructed to locate and read trd.md and config.toml in references/discovering-supabase-root.md to determine infrastructure and environment settings.
Boundary markers: No specific delimiters or safety warnings are provided for the agent when processing the content of these files.
Capability inventory: The skill has the ability to execute shell commands, write workflow files, and set repository secrets via the GitHub CLI.
Sanitization: There are no explicit sanitization routines defined for values extracted from project files before they are interpolated into CLI commands.
[CREDENTIALS_SAFE]: The skill includes explicit instructions for the secure management of credentials, ensuring they are handled correctly and not leaked.
Evidence: SKILL.md requires checking that *.secrets.env files are ignored by git (git check-ignore). references/configuring-remote-project.md uses a subshell pattern ( set -a; . <file>; ... ) to load environment variables without leaking them to the parent shell, and explicitly advises against printing secret values.

deploying-supabase-application