Gemini

• You want to query Google Gemini for a quick answer, summary, or generated text without setting up a full API integration.
• You need structured output (e.g., JSON) from a large language model as part of an automated pipeline.
• You want to switch between Gemini model variants using a simple --model flag.
• You are already using the Gemini CLI and want your agent to leverage the same toolchain.

• You need a multi-turn, interactive conversation — this skill is designed for one-shot prompts only.
• You prefer OpenAI or Anthropic models; use a skill backed by those providers instead.

Submit a prompt as a positional argument and receive an immediate response. Designed to avoid interactive mode, making it reliable for automated agent use.

Choose any available Gemini model variant at runtime using the --model <name> flag, giving you flexibility over quality, speed, and cost.

Use --output-format json to instruct Gemini to return machine-readable JSON, ideal for pipelines that need to parse or forward the result.

List and manage Gemini CLI extensions with built-in commands (--list-extensions, gemini extensions <command>), allowing capability expansion over time.

Ask Gemini a direct question and get an answer in seconds — useful for agents that need to resolve a factual query mid-workflow without human intervention.

Pass a block of text as a prompt and ask Gemini to summarize it, enabling automated document or article condensation in a pipeline.

Request JSON-formatted output from Gemini to feed downstream tools, such as generating a structured report or extracting entities from a description.

Generate drafts, descriptions, or creative content by supplying a prompt, letting Gemini handle the writing while your agent handles orchestration.