The Agent Skills Directory

[EXTERNAL_DOWNLOADS]: The skill references the official NVIDIA Fleet Intelligence client releases on GitHub (https://github.com/NVIDIA/fleet-intelligence-client/releases) for installation. This is a trusted resource provided by the vendor.
[CREDENTIALS_SAFE]: The skill contains strong defensive instructions regarding credential management. It explicitly forbids the agent from requesting, printing, logging, or inputting API keys on behalf of the user. It directs users to use interactive CLI prompts or official NVIDIA service key pages for authentication setup.
[COMMAND_EXECUTION]: The skill utilizes the nvfleetint CLI to perform fleet health queries. The use of this tool is the primary intended function and is restricted to pre-defined query commands with JSON output formatting for safer parsing.
[INDIRECT_PROMPT_INJECTION]: The skill processes data returned by the nvfleetint CLI to answer user questions.
Ingestion points: Data enters the context via the standard output of the nvfleetint command.
Boundary markers: The instructions mandate the use of live JSON data but do not specify explicit delimiters for the injected content.
Capability inventory: The skill is restricted to executing the nvfleetint CLI tool.
Sanitization: The skill emphasizes providing prose answers based on the structured JSON output rather than raw data passthrough, reducing the risk of content-based manipulation.

nvfleetint