The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill ingests and processes data from PostgreSQL query results, which may include untrusted content from the database environment. * Ingestion points: SQL query results from system views (e.g., pg_stat_activity, pg_stat_statements) and user-defined table data analyzed during query optimization. * Boundary markers: The instructions do not specify delimiters or guidelines for the agent to distinguish between its instructions and potentially malicious instructions embedded in database records. * Capability inventory: The skill is designed to execute high-privilege administrative SQL commands, including ALTER SYSTEM, CREATE INDEX, and table partitioning. * Sanitization: There are no explicit instructions for the agent to sanitize or validate data retrieved from the database before interpreting it.

postgres-expert