The Agent Skills Directory

[SAFE]: The skill is entirely defensive in nature, providing instructions for an AI agent to act as a security linter for PostgreSQL Row Level Security implementations.
[COMMAND_EXECUTION]: The skill suggests using the Bash tool to perform local development tasks such as identifying modified files (git diff) and running database tests (psql). These commands are standard for the stated purpose of the skill and do not involve network exfiltration or unauthorized access.
[INDIRECT_PROMPT_INJECTION]: The skill instructs the agent to analyze untrusted data (project migration files and source code). While this presents an attack surface, the skill's logic is focused on pattern matching for known vulnerabilities (e.g., missing FORCE ROW LEVEL SECURITY) and does not incorporate data into its own command execution or prompt logic in an unsafe manner. The capability is used for security verification.

postgres-rls