PluginBench
Skill
Pass
Audit score 90

gemini

steipete/clawdis

Gemini CLI for one-shot prompts, summaries, generation, and routing via skills, hooks, or MCP.

What is gemini?

Gemini CLI enables headless, one-shot interactions with Google's Gemini models. Use it for quick prompts, content generation, summaries, and extensible workflows through skills, hooks, and MCP integrations.

  • Execute one-shot prompts with `-p/--prompt` flag
  • Pipe stdin to append context to prompts
  • Select models with `-m/--model` and format output as JSON or other formats
  • Manage extensions, skills, hooks, and MCP integrations
  • Route requests through Gemma or other configured models

How to install gemini

npx skills add https://github.com/steipete/clawdis --skill gemini
Prerequisites
  • Gemini CLI binary installed (via Homebrew: `brew install gemini-cli`)
  • Google authentication configured (run `gemini` interactively once to log in)
Claude Code
Cursor
Windsurf
Cline

How to use gemini

  1. 1.Install Gemini CLI using Homebrew or your package manager
  2. 2.Run `gemini` once interactively to authenticate with Google
  3. 3.Use `gemini -p "your prompt"` for one-shot queries
  4. 4.Pipe input with `cat file.md | gemini -p "Summarize"`
  5. 5.Explore extensions with `gemini --list-extensions` and manage with `gemini extensions <command>`
  6. 6.Configure skills, hooks, and MCP as needed for your workflow

Use cases

Good for
  • Summarize documents or notes from the command line
  • Generate content or answers in batch or scripted workflows
  • Integrate Gemini into shell pipelines for text processing
  • Extend functionality with custom skills and hooks
  • Connect to external tools via MCP for enhanced capabilities
Who it's for
  • CLI-first developers and engineers
  • DevOps and automation specialists
  • Content creators working in terminal environments
  • Teams building AI-powered workflows and integrations

gemini FAQ

How do I authenticate with Gemini CLI?

Run `gemini` once interactively and follow the login flow to set up Google authentication.

Can I use Gemini CLI in scripts and pipelines?

Yes, use the `-p/--prompt` flag for headless one-shot mode and pipe stdin to append context.

How do I select a specific model?

Use the `-m/--model` flag to specify which model to use, e.g., `gemini -m <model> -p "Prompt"`

What output formats are supported?

Use `--output-format json` for JSON output, or other formats depending on your needs.

What is the `--yolo` flag and should I use it?

The `--yolo` flag bypasses safety checks; avoid it for security and safety reasons.

Full instructions (SKILL.md)

Source of truth, from steipete/clawdis.


name: gemini description: "Gemini CLI one-shot prompts, summaries, generation, skills, hooks, MCP, or Gemma routing." homepage: https://ai.google.dev/ metadata: { "openclaw": { "emoji": "✨", "requires": { "bins": ["gemini"] }, "install": [ { "id": "brew", "kind": "brew", "formula": "gemini-cli", "bins": ["gemini"], "label": "Install Gemini CLI (brew)", }, ], }, }

Gemini CLI

Use Gemini in headless one-shot mode. Positional text starts interactive mode; use -p/--prompt.

Quick start

  • gemini -p "Answer this question..."
  • gemini -m <model> -p "Prompt..."
  • gemini -p "Return JSON" --output-format json
  • stdin appends to -p: cat notes.md | gemini -p "Summarize"

Extensions

  • List: gemini --list-extensions
  • Manage: gemini extensions <command>
  • Skills: gemini skills <command>
  • Hooks: gemini hooks <command>
  • MCP: gemini mcp <command>

Notes

  • If auth is required, run gemini once interactively and follow the login flow.
  • Avoid --yolo for safety.