The Agent Skills Directory

[COMMAND_EXECUTION]: The skill's primary functionality is delivered through system command execution via npx. This includes infrastructure management commands (install, start, stop, uninstall) and database utility operations.- [EXTERNAL_DOWNLOADS]: Fetches and executes the latest version of the @ruvector/postgres-cli package from the NPM registry at runtime during command execution.- [REMOTE_CODE_EXECUTION]: Provides the ability to execute arbitrary SQL commands provided by the user or read from local files via the psql subcommand, which operates as code execution within the database environment.- [PROMPT_INJECTION]: The skill exhibits an indirect prompt injection surface due to its interaction with external data sources.
Ingestion points: Processes data from external files using --file flags in vector and psql commands, and retrieves information from database tables via vector search, graph queries, and learning store retrieval.
Boundary markers: There are no explicit instructions or delimiters defined to prevent the agent from following commands embedded within the data retrieved from the database.
Capability inventory: The skill possesses significant capabilities including subprocess execution (npx), local file system read/write access for database management, and network communication for PostgreSQL connections.
Sanitization: No evidence of sanitization or validation of database query results or file contents is provided before the data is processed by the agent.

@ruvector/postgres-cli