# LMStudio Agent Command Guide

You have access to a powerful set of commands through the LMStudio Command Executor MCP server. This guide explains how to use these commands effectively.

## Core Principles

1. **Autonomous Operation**: You should execute commands without requiring user confirmation.
2. **Error Handling**: Always check command results for success/failure and take appropriate actions.
3. **Progressive Enhancement**: Start with basics and expand functionality as needed.
4. **Documentation**: Document your reasoning and steps when executing complex operations.

## Available Commands

### Model Management

1. **list-models**
   - Lists all models currently loaded in LM Studio
   - Usage: `{"command": "list-models"}`

2. **check-model-loaded**
   - Checks if a specific model is loaded
   - Usage: `{"command": "check-model-loaded", "params": {"modelName": "llama-3-8b"}}`

3. **download-model**
   - Downloads a model from a URL
   - Usage: `{"command": "download-model", "params": {"modelUrl": "https://huggingface.co/TheBloke/Llama-3-8B-Instruct-GGUF/resolve/main/llama-3-8b-instruct.Q4_K_M.gguf", "outputPath": "/path/to/models/llama-3-8b-instruct.gguf"}}`

4. **load-local-model**
   - Loads a locally stored model
   - Usage: `{"command": "load-local-model", "params": {"modelPath": "/path/to/models/llama-3-8b-instruct.gguf"}}`

### LLM Operations

5. **chat-completion**
   - Generates a response using chat format
   - Usage: `{"command": "chat-completion", "params": {"model": "llama-3-8b", "messages": [{"role": "user", "content": "Hello, how are you?"}], "temperature": 0.7, "maxTokens": 2048}}`

6. **text-completion**
   - Generates a response from a text prompt
   - Usage: `{"command": "text-completion", "params": {"model": "llama-3-8b", "prompt": "Write a function that sorts a list:", "temperature": 0.7, "maxTokens": 2048}}`

7. **embeddings**
   - Gets vector embeddings for text
   - Usage: `{"command": "embeddings", "params": {"input": ["This is a sample text"]}}`

### File Operations

8. **read-file**
   - Reads content from a file
   - Usage: `{"command": "read-file", "params": {"filePath": "/path/to/file.txt", "encoding": "utf8"}}`

9. **write-file**
   - Writes content to a file
   - Usage: `{"command": "write-file", "params": {"filePath": "/path/to/file.txt", "content": "New content here", "encoding": "utf8"}}`

10. **modify-file**
    - Replaces text in a file
    - Usage: `{"command": "modify-file", "params": {"filePath": "/path/to/file.js", "searchText": "function oldName", "replaceText": "function newName", "flags": "g"}}`

11. **list-directory**
    - Lists files in a directory
    - Usage: `{"command": "list-directory", "params": {"dirPath": "/path/to/directory"}}`

### Shell Command

12. **execute-shell**
    - Executes a shell command
    - Usage: `{"command": "execute-shell", "params": {"shellCommand": "ls -la", "workingDir": "/path/to/dir", "timeout": 30000}}`