A bridge for utilizing the local Gemini CLI to perform secondary model reasoning, content generation, and structured data tasks on Linux hosts.
Install & Download
1. ClawHub CLI
The fastest way to install a skill directly from the registry.
npx clawhub@latest install openclaw-gemini-cli
2. Manual Installation
Copy the skill folder to one of these locations
~/.openclaw/skills/ <project>/skills/ Priority: Workspace > Local > Bundled
3. Prompt Installation
Copy this prompt to OpenClaw to install it automatically.
Help me install openclaw-gemini-cli using Clawhub. If Clawhub is not installed, install it first (npm i -g clawhub).
Prefer to download?
Get the raw skill files in a ZIP archive.
What is Gemini CLI Integration?
The Gemini CLI skill empowers Openclaw Skills to utilize a locally installed Gemini binary as a secondary reasoning engine. This integration is designed for developers who need one-shot prompts, structured JSON output, or shell-assisted research directly within their Linux environment. Instead of relying solely on a single model, agents can delegate specific drafting, naming, or code-review tasks to Gemini while maintaining full control over the file system and command execution.
This tool is optimized for high-performance, low-latency interactions where the agent acts as the primary orchestrator. By leveraging the gemini command, users can ask for second opinions on complex logic or generate production-ready copy without leaving the CLI. It bridges the gap between autonomous file editing and sophisticated model-driven planning, making it a cornerstone for advanced Openclaw Skills implementations.
Gemini CLI Integration Use Cases
- Obtaining a second opinion from a different model during complex architectural decisions
- Drafting UI text, page copy, SQL schemas, or shell scripts for local review
- Generating strict JSON payloads for automated post-processing by other scripts
- Managing AI-to-AI workflows where Openclaw manages files while Gemini generates the implementation plan
- Running rapid one-shot prompts for summarization or research without spawning new sessions
How Gemini CLI Integration Works
- The agent inspects the local environment to verify Gemini CLI availability and authentication status.
- Openclaw Skills prepares a bounded prompt, often incorporating local file context through shell piping or redirection.
- The system executes the gemini command with specific flags to capture output as raw text or structured data.
- The agent reviews the generated output against safety boundaries to ensure no destructive shell commands are piped into execution.
- Depending on the workflow, the output is either displayed to the user, saved to a temporary file, or used to modify the local codebase.
Gemini CLI Integration Setup
Ensure the gemini binary is installed and available in your PATH. You must complete the authentication process before utilizing this skill within Openclaw Skills.
# Verify installation
which gemini
# Run a smoke test to verify auth
gemini -p "hello"
# If auth fails, follow the CLI login prompts
Note: On some Linux distributions, you may see warnings regarding missing keytar dependencies; the skill will automatically fall back to FileKeychain and remain functional.
Gemini CLI Integration Data Schema & Taxonomy
| Data Component | Logic/Pattern | Purpose |
|---|---|---|
| One-shot Output | gemini -p "prompt" |
Quick text generation and analysis. |
| JSON Capture | gemini -p "JSON prompt" > /tmp/out.json |
Programmatic data handling for automation. |
| File Context | `cat |
gemini -p "$(cat)"` |
| Model Selection | gemini --help |
Version-specific model configuration and flags. |
| Helper Scripts | scripts/gemini_*.sh |
Pre-defined wrappers for review and JSON tasks. |
Gemini CLI Integration Advanced Features
- Multi-step AI-to-AI delegation: Orchestrate workflows where one model plans and Openclaw Skills executes
- Strict JSON enforcement: Use specialized prompts to receive schema-validated data for integration pipelines
- Large prompt management: Utilize temporary files and shell substitution to handle complex context beyond standard CLI limits
- Model switching: Explicitly define model versions via CLI flags to ensure deterministic output for critical tasks
- Automated review loops: Integrate the gemini_review script to perform recurring quality checks on local file changes
SKILL.md
Loading
Related Openclaw Skills
Lawyer Six Steps
A professional legal assistant skill implementing the Chinese Lawyer Six-Step Method to generate comprehensive case analysis reports.
Team Tasks
A sophisticated coordination framework for managing multi-agent development pipelines through shared JSON task files and centralized AGI dispatch.
n8n Workflow Management
A comprehensive management tool for n8n to programmatically create, test, and optimize automated workflows.
Loyalty Designer
A strategic framework for designing margin-positive loyalty programs that increase customer lifetime value through points, tiers, and referral mechanics.
Session OpenClaw Bridge
A specialized bridge skill for connecting Session Messenger with n8n workflows and OpenClaw agents for secure, automated messaging.
Hair Salon Promo Video
A specialized AI video agent designed to turn salon transformations and stylist expertise into high-conversion social media marketing content.
Featured*
