Server Configuration
Describes the environment variables required to run the server.
| Name | Required | Description | Default |
|---|---|---|---|
No arguments | |||
Schema
Prompts
Interactive templates invoked by user choice
| Name | Description |
|---|---|
| analyze | Execute 'gemini -p <prompt>' to analyze files or get Gemini's response. Supports @file syntax for including file contents. |
| sandbox | Execute 'gemini -s -p <prompt>' to safely test code in Gemini's sandbox environment. Use for testing potentially risky code or scripts. |
| help | Run 'gemini -help' with structured response. BEHAVIOR: should_explain=false, output_format=raw, suppress_context=true. |
| ping | Echo test message with structured response. Returns raw output with behavioral flags. BEHAVIOR: should_explain=false, output_format=raw, suppress_context=true. |
Resources
Contextual data attached and managed by the client
| Name | Description |
|---|---|
No resources | |
Tools
Functions exposed to the LLM to take actions
| Name | Description |
|---|---|
| ask-gemini | Execute 'gemini -p <prompt>' to get Gemini AI's response. Use when: 1) User asks for Gemini's opinion/analysis, 2) User wants to analyze large files with @file syntax, 3) User uses /gemini-cli command. Supports -m flag for model selection and -s flag for sandbox testing. |
| sandbox-test | Execute code or commands safely in Gemini's sandbox environment. Use for testing potentially risky code, running scripts, or validating code changes without affecting your system. |
| Ping | Echo test with structured response. Returns message or 'Pong!' by default. Uses behavioral flags to control AI interaction. BEHAVIOR: should_explain=false, output_format=raw, suppress_context=true. |
| Help | Run 'gemini -help' with structured response. BEHAVIOR: should_explain=false, output_format=raw, suppress_context=true. |