### Quick Start Commands

Source: https://github.com/aaif-goose/goose/blob/main/evals/open-model-gym/README.md

Provides essential commands for setting up and running the Open Model Gym. Use 'just install' for one-time setup, 'just run' to execute the full test matrix, and 'just report' to view results.

```bash
just install   # one-time setup
just run       # run full matrix (3 reps each)
just report    # view results
```

--------------------------------

### Example Output of Plugin Installation

Source: https://github.com/aaif-goose/goose/blob/main/documentation/docs/guides/context-engineering/plugins.md

This is an example of the output you can expect after successfully installing a plugin. It shows the plugin name, version, source, location, and imported skills.

```text
✓ Installed open-plugins plugin 'my-plugin' (1.0.0)
  Source: https://github.com/example/my-goose-plugin.git
  Location: /Users/you/.agents/plugins/my-plugin
  Imported skills:
    - my-plugin:review
    - my-plugin:test-plan
```

--------------------------------

### Install and Run Example Hook Plugin

Source: https://github.com/aaif-goose/goose/blob/main/documentation/docs/guides/context-engineering/hooks.md

This snippet shows how to install the example 'hello-hooks' plugin and run a Goose session to observe its behavior. The plugin prints hook events to stderr and logs full payloads to a file.

```bash
mkdir -p ~/.agents/plugins
cp -R examples/plugins/hello-hooks ~/.agents/plugins/hello-hooks
chmod +x ~/.agents/plugins/hello-hooks/scripts/announce.sh

goose session
```

--------------------------------

### Install MLflow and Start Server

Source: https://github.com/aaif-goose/goose/blob/main/documentation/docs/tutorials/mlflow.md

Install the MLflow Python package and start the MLflow tracking server. The UI will be accessible at http://localhost:5000.

```bash
pip install mlflow
mlflow server --port 5000
```

--------------------------------

### Enable Example Streamable HTTP Extension

Source: https://github.com/aaif-goose/goose/blob/main/documentation/docs/getting-started/using-extensions.md

Example of starting a session with a specific Streamable HTTP extension by providing its example URL.

```bash
goose session --with-streamable-http-extension "https://example.com/streamable"
```

--------------------------------

### Start New Session for API Development

Source: https://github.com/aaif-goose/goose/blob/main/documentation/docs/guides/managing-projects.md

This example shows how to change to a project directory and start a new named session for API development.

```bash
cd ~/projects/ecommerce-api
goose session --name "api-auth-work"
```

--------------------------------

### Workflow Setup Subrecipe Output Example

Source: https://github.com/aaif-goose/goose/blob/main/documentation/blog/2025-07-28-streamlining-detection-development-with-goose-recipes/index.md

Provides a JSON example of the output from the workflow_setup subrecipe, detailing the structure of repository readiness, branch information, and Panther MCP access test results.

```json
{
  "status": { "success": true },
  "data": {
    "branch_name": "ai/okta-suspicious-login",
    "standards_summary": "Rules must use ai_ prefix, implement required functions...",
    "repo_ready": true,
    "mcp_panther": { "access_test_successful": true }
  }
}
```

--------------------------------

### Example Skill with Supporting Files

Source: https://github.com/aaif-goose/goose/blob/main/documentation/docs/guides/context-engineering/using-skills.md

Demonstrates a skill that includes a bash script and a JSON template for API setup.

```markdown
---
name: api-setup
description: Set up API integration with configuration and helper scripts
---

# API Setup

This skill helps you set up a new API integration with our standard configuration.

## Steps

1. Run `setup.sh <api-name>` to create the integration directory
2. Copy `templates/config.template.json` to your integration directory
3. Update the config with your API credentials
4. Test the connection

## Configuration

The config template includes:
- `api_key`: Your API key (get from the provider's dashboard)
- `endpoint`: API endpoint URL
- `timeout`: Request timeout in seconds (default: 30)

## Verification

After setup, verify:
- [ ] Config file is valid JSON
- [ ] API key is set and not a placeholder
- [ ] Test connection succeeds

```

```bash
#!/bin/bash
API_NAME=$1
mkdir -p "integrations/$API_NAME"
cp templates/config.template.json "integrations/$API_NAME/config.json"
echo "Created integration directory for $API_NAME"
echo "Edit integrations/$API_NAME/config.json with your credentials"

```

```json
{
  "api_key": "YOUR_API_KEY_HERE",
  "endpoint": "https://api.example.com/v1",
  "timeout": 30,
  "retry_attempts": 3
}

```

--------------------------------

### Example Allowlist Configuration

Source: https://github.com/aaif-goose/goose/blob/main/documentation/docs/guides/allowlist.md

An example YAML configuration specifying allowed extensions for Slack, GitHub, and Jira. Only these extensions, identified by their commands, can be installed.

```yaml
extensions:
  - id: slack
    command: uvx mcp_slack
  - id: github
    command: uvx mcp_github
  - id: jira
    command: uvx mcp_jira
```

--------------------------------

### Setup Hermit Environment and Build Project

Source: https://github.com/aaif-goose/goose/blob/main/AGENTS.md

Source the environment activation script and build the project. This is a common starting point for development.

```bash
source bin/activate-hermit
cargo build
```

--------------------------------

### Install Dev.to MCP Server

Source: https://github.com/aaif-goose/goose/blob/main/documentation/docs/mcp/dev.to-mcp.md

Clone the Dev.to MCP server repository, install dependencies, build the project, and start the server. Ensure Node.js and npm are installed.

```bash
git clone https://github.com/nickytonline/dev-to-mcp.git
cd dev-to-mcp
npm install
npm run build # one-time setup
```

```bash
npm start
```

--------------------------------

### Start Specific Tutorial by Name

Source: https://github.com/aaif-goose/goose/blob/main/documentation/docs/mcp/tutorial-mcp.md

Directly request a tutorial by its name, for example, asking goose to walk through the 'first-game' tutorial.

```text
( O)> Can you walk me through the first-game tutorial?
```

--------------------------------

### Start New Session for Admin Dashboard

Source: https://github.com/aaif-goose/goose/blob/main/documentation/docs/guides/managing-projects.md

This example shows how to change to a project directory and start a new named session for working on an admin dashboard.

```bash
cd ~/projects/admin-dashboard
goose session --name "dashboard-ui"
```

--------------------------------

### Start and Configure Provider Error Proxy

Source: https://github.com/aaif-goose/goose/blob/main/scripts/provider-error-proxy/README.md

This example shows how to start the proxy and configure your application (Goose) to use it by setting environment variables. It also displays the initial status of the proxy.

```bash
$ uv run proxy.py
============================================================
🔧 Provider Error Proxy
============================================================
Port: 8888

To use with Goose, set these environment variables:
  export OPENAI_HOST=http://localhost:8888
  export ANTHROPIC_HOST=http://localhost:8888
  ...
============================================================

============================================================
Current mode: ✅ No error (pass through)
Requests handled: 0
============================================================

Commands:
  n      - No error (pass through) - permanent
  c      - Context length exceeded (1 time)
  c 4    - Context length exceeded (4 times)
  c 0.3  - Context length exceeded (30% of requests)
  c 30%  - Context length exceeded (30% of requests)
  c *    - Context length exceeded (100% of requests)
  r      - Rate limit error (1 time)
  u      - Unknown server error (1 time)
  q      - Quit

Enter command: r
============================================================
Current mode: ⏱️  Rate limit exceeded (1 remaining)
Requests handled: 0
============================================================
...
2025-10-09 14:30:15 - __main__ - INFO - 📨 Request #1: POST /v1/chat/completions -> openai
2025-10-09 14:30:15 - __main__ - WARNING - 💥 Injecting RATE_LIMIT error (status 429) for openai

Enter command: n
============================================================
Current mode: ✅ No error (pass through)
Requests handled: 1
============================================================
...
2025-10-09 14:30:20 - __main__ - INFO - 📨 Request #2: POST /v1/chat/completions -> openai
2025-10-09 14:30:20 - __main__ - INFO - ✅ Proxied response: 200

```

--------------------------------

### Troubleshoot Beads MCP Extension Installation

Source: https://github.com/aaif-goose/goose/blob/main/documentation/docs/mcp/beads-mcp.md

If the extension fails to start due to a missing `packaging` module, install it explicitly using `uv tool install`.

```sh
uv tool install beads-mcp --with packaging
```

--------------------------------

### Install container-use on all platforms

Source: https://github.com/aaif-goose/goose/blob/main/documentation/blog/2025-06-19-isolated-development-environments/index.md

Use this command to install the container-use CLI on any platform by downloading and executing the install script.

```bash
curl -fsSL https://raw.githubusercontent.com/dagger/container-use/main/install.sh | bash
```

--------------------------------

### Goose CLI Configuration Flow Example

Source: https://github.com/aaif-goose/goose/blob/main/documentation/docs/getting-started/installation.md

Illustrates the interactive command-line interface prompts and feedback during the setup of a Tetrate Agent Router Service provider.

```bash
┌   goose-configure
│
◇ How would you like to set up your provider?
│ Tetrate Agent Router Service Login
│
Opening browser for Tetrate Agent Router Service authentication...
[goose opens the browser and prints details]

Authentication complete!

Configuring Tetrate Agent Router Service...
✓ Tetrate Agent Router Service configuration complete
✓ Models configured successfully

Testing configuration...
✓ Configuration test passed!
✓ Developer extension enabled!
└ Tetrate Agent Router Service setup complete! You can now use goose.
```

--------------------------------

### Start New Session for Mobile App Bug Fix

Source: https://github.com/aaif-goose/goose/blob/main/documentation/docs/guides/managing-projects.md

This example demonstrates navigating to a project directory and starting a new session to debug a mobile application.

```bash
cd ~/projects/mobile-app
goose session
```

--------------------------------

### Install Fetch Extension via CLI

Source: https://github.com/aaif-goose/goose/blob/main/documentation/docs/mcp/fetch-mcp.md

Use the `uvx` command to quickly install the Fetch MCP Server extension. Ensure `uv` is installed on your system.

```sh
uvx mcp-server-fetch
```

--------------------------------

### Install Dependencies

Source: https://github.com/aaif-goose/goose/blob/main/documentation/docs/tutorials/subagents.md

Navigate to your project directory and install the necessary Node.js dependencies.

```bash
cd your-project-folder
npm install
```

--------------------------------

### Start Documentation Development Server

Source: https://github.com/aaif-goose/goose/blob/main/documentation/blog/README.md

Navigate to the documentation directory and start the development server to preview blog posts locally.

```bash
cd documentation
npm start
```

--------------------------------

### Install goose CLI

Source: https://github.com/aaif-goose/goose/blob/main/README.md

Installs the goose command-line interface using a script. This is a quick way to get started with goose on your terminal.

```bash
curl -fsSL https://github.com/aaif-goose/goose/releases/download/stable/download_cli.sh | bash
```

--------------------------------

### Install Documentation Dependencies

Source: https://github.com/aaif-goose/goose/blob/main/documentation/blog/README.md

Install the necessary Node.js dependencies for the documentation site.

```bash
cd documentation
npm install
```

--------------------------------

### Select Configuration Option

Source: https://github.com/aaif-goose/goose/blob/main/documentation/docs/getting-started/providers.md

Choose 'Configure Providers' from the main configuration menu to start.

```text
┌   goose-configure 
│
◆  What would you like to configure?
│  ● Configure Providers (Change provider or update credentials)
│  ○ Custom Providers 
│  ○ Add Extension 
│  ○ Toggle Extensions 
│  ○ Remove Extension 
│  ○ goose Settings 
└  
```

--------------------------------

### Start Building with RPI

Source: https://github.com/aaif-goose/goose/blob/main/documentation/blog/2026-02-06-rpi-openclaw-alternative/index.md

Use this command to initiate the building process based on a pre-defined plan. Ensure the plan is ready before executing.

```bash
/implement_plan start building
```

--------------------------------

### Example goose Prompt

Source: https://github.com/aaif-goose/goose/blob/main/documentation/docs/experimental/vs-code-extension.md

This is an example prompt to instruct goose to update a contributing guide with specific instructions.

```plaintext
Update the contributing guide at /goose/CONTRIBUTING.md with instructions on how to start docusaurus dev server
```

--------------------------------

### Start Tutorial via Natural Language Prompt

Source: https://github.com/aaif-goose/goose/blob/main/documentation/docs/mcp/tutorial-mcp.md

Initiate a tutorial by asking goose directly about a topic of interest. Goose will suggest a relevant tutorial, such as 'build-mcp-extension' for learning extension development.

```text
( O)> I'd like to learn how to build an extension for goose
```

```text
I'll help you learn about building extensions! The "build-mcp-extension" tutorial would be perfect for this. Let me load it and guide you through it.

─── load_tutorial | tutorial ──────────────────────────
name: build-mcp-extension



Before we dive into building your extension, I need to understand a few things to better guide you:

1. Which SDK would you prefer to use? We have three options:
   - Python (beginner-friendly, great for prototyping)
   - TypeScript (great for web-based tools)
   - Kotlin (robust, type-safe, good for complex applications)

2. What type of extension would you like to build? For example:
   - A tool that performs specific actions
   - A resource that provides data to the LLM
   - Or do you have a specific idea in mind?

Let me know your preferences, and I'll help you get started with the appropriate setup and implementation!

For reference, here's what we'll cover:
1. Setting up your development environment
2. Creating the basic MCP server
3. Implementing your specific functionality
4. Testing and debugging

Which SDK would you like to use to get started?
```

--------------------------------

### Install Alby MCP Extension (Desktop UI)

Source: https://github.com/aaif-goose/goose/blob/main/documentation/docs/mcp/alby-mcp.md

Launch the installer for the Alby MCP extension directly from the goose Desktop application. This will guide you through the installation process.

```link
goose://extension?cmd=npx&arg=-y&arg=%40getalby%2Fmcp&id=alby&name=Alby&description=Connect%20goose%20to%20your%20Bitcoin%20Lightning%20Wallet&env=NWC_CONNECTION_STRING%3DNWC%20Connection%20Secret
```

--------------------------------

### Set up Goose Plugin Example

Source: https://github.com/aaif-goose/goose/blob/main/documentation/blog/2026-05-14-goose-hooks/index.md

Copies a sample Goose plugin to the local plugins directory and makes its script executable. This prepares the environment to test hook functionality.

```bash
mkdir -p ~/.agents/plugins
cp -R examples/plugins/hello-hooks ~/.agents/plugins/hello-hooks
chmod +x ~/.agents/plugins/hello-hooks/scripts/announce.sh
```

--------------------------------

### Install and Run goose ACP TUI

Source: https://github.com/aaif-goose/goose/blob/main/ui/text/README.md

Install dependencies and start the TUI using pnpm. This runs against the released binary.

```bash
cd ui/text
pnpm install   # pulls the pinned @aaif/goose-sdk and its matching @aaif/goose-binary-* package
pnpm start     # tsx src/tui.tsx — runs against the released binary, no Rust build
```

--------------------------------

### Set up Project Environment

Source: https://github.com/aaif-goose/goose/blob/main/documentation/docs/tutorials/custom-extensions.md

Installs project dependencies. Ensure Node.js and npm are installed for MCP Inspector.

```bash
uv sync
```

--------------------------------

### Install Figma Extension via Goose Desktop

Source: https://github.com/aaif-goose/goose/blob/main/documentation/docs/mcp/figma-mcp.md

Use this link to launch the Goose Desktop installer for the Figma extension. This allows for quick setup and integration.

```html
[Launch the installer](goose://extension?type=streamable_http&url=http%3A%2F%2F127.0.0.1%3A3845%2Fmcp&id=figma&name=Figma&description=Convert%20Figma%20designs%20into%20code%20and%20extract%20design%20context)
```

--------------------------------

### Install AgentQL MCP Extension via Goose Desktop

Source: https://github.com/aaif-goose/goose/blob/main/documentation/docs/mcp/agentql-mcp.md

Use this link to launch the installer for the AgentQL MCP extension directly within the Goose Desktop application. It includes installation command and environment variable setup.

```html
[Launch the installer](goose://extension?cmd=npx&arg=-y&arg=agentql-mcp&id=agentql&name=AgentQL&description=Transform%20unstructured%20web%20content%20into%20structured%20data&env=AGENTQL_API_KEY%3DAgentQL%20API%20Key)
```

--------------------------------

### Install and Run hello-hooks Plugin

Source: https://github.com/aaif-goose/goose/blob/main/examples/plugins/hello-hooks/README.md

Steps to install the plugin and observe its output during a Goose session. Inspect the log file to see the full event payloads.

```bash
mkdir -p ~/.agents/plugins
cp -R examples/plugins/hello-hooks ~/.agents/plugins/hello-hooks
chmod +x ~/.agents/plugins/hello-hooks/scripts/announce.sh

# Then run goose normally; you should see lines like
# 🚀 [hello-hooks] SessionStart
# 💬 [hello-hooks] UserPromptSubmit
# ⚡ [hello-hooks] PreToolUse tool=developer__shell
# ✅ [hello-hooks] PostToolUse tool=developer__shell
goose session

# Inspect the full payloads goose passed to the hook:
tail ~/.agents/plugins/hello-hooks/last-event.log
```

--------------------------------

### Initialize Node.js Project

Source: https://github.com/aaif-goose/goose/blob/main/documentation/docs/tutorials/building-mcp-apps.md

Sets up a new directory for your project and initializes a Node.js project with default settings.

```bash
mkdir mcp-app-demo
cd mcp-app-demo
npm init -y
```

--------------------------------

### Launch Neon Remote MCP Installer via Goose Desktop

Source: https://github.com/aaif-goose/goose/blob/main/documentation/docs/mcp/neon-mcp.md

Use this URL to launch the installer for the Neon Remote MCP extension directly from Goose Desktop. This simplifies the setup process.

```link
goose://extension?type=streamable_http&url=https%3A%2F%2Fmcp.neon.tech%2Fmcp&id=neon&name=Neon&description=Manage%20Neon%20Postgres%20databases%2C%20projects%2C%20and%20branches
```

--------------------------------

### Initialize Project and Install Dependencies

Source: https://github.com/aaif-goose/goose/blob/main/documentation/docs/tutorials/remotion-video-creation.md

These shell commands initialize a new project directory for the git terminal video and install the necessary Remotion and React dependencies.

```bash
mkdir -p git-terminal-video && cd git-terminal-video && npm init -y
```

```bash
npm install remotion @remotion/cli react react-dom typescript @types/react
```

--------------------------------

### Initialize MCP Project with uv

Source: https://github.com/aaif-goose/goose/blob/main/documentation/docs/tutorials/custom-extensions.md

Sets up a new Python project for an MCP server using uv. Creates a basic directory structure and necessary files.

```bash
uv init --lib mcp-wiki
cd mcp-wiki

mkdir -p src/mcp_wiki
touch src/mcp_wiki/server.py
touch src/mcp_wiki/__main__.py
```

--------------------------------

### Goose Plugin Commands Usage

Source: https://github.com/aaif-goose/goose/blob/main/documentation/docs/guides/goose-cli-commands.md

Examples for installing and updating plugins using the 'goose plugin' command.

```bash
goose plugin install https://github.com/example/my-goose-plugin.git
```

```bash
goose plugin install --auto-update https://github.com/example/my-goose-plugin.git
```

```bash
goose plugin update my-plugin
```

--------------------------------

### Launch Neon Local MCP Installer via Goose Desktop

Source: https://github.com/aaif-goose/goose/blob/main/documentation/docs/mcp/neon-mcp.md

Use this URL to launch the installer for the Neon Local MCP extension from Goose Desktop. It includes the command to start the server with your API key.

```link
goose://extension?cmd=npx&arg=-y&arg=%40neondatabase%2Fmcp-server-neon&arg=start&arg=%3CYOUR_NEON_API_KEY%3E&id=neon&name=Neon&description=Manage%20your%20Neon%20Postgres%20databases%2C%20projects%2C%20and%20branches
```

--------------------------------

### Example Settings Configuration (OpenAI)

Source: https://github.com/aaif-goose/goose/blob/main/documentation/docs/guides/recipes/recipe-reference.md

Configure AI provider and model settings for a recipe, overriding defaults.

```yaml
settings:
  goose_provider: "openai"
  goose_model: "gpt-4o"
  temperature: 0.3
```

--------------------------------

### Example Slash Command Usage

Source: https://github.com/aaif-goose/goose/blob/main/documentation/docs/guides/context-engineering/slash-commands.md

Type a custom command with a leading slash at the start of your message to invoke a recipe.

```bash
/daily-report
```

--------------------------------

### Start Local Development Server

Source: https://github.com/aaif-goose/goose/blob/main/documentation/README.md

Starts a local development server and opens the website in a browser. Changes are reflected live without server restarts.

```bash
npm run start
```

--------------------------------

### Try the new Goose TUI

Source: https://github.com/aaif-goose/goose/blob/main/documentation/blog/2026-04-08-goose-acp-and-new-tui/index.md

Run this command to start an interactive session with the latest beta of the TypeScript-based TUI. No installation is required.

```bash
npx @aaif/goose
```

--------------------------------

### Install and Build MCP Harness

Source: https://github.com/aaif-goose/goose/blob/main/evals/open-model-gym/README.md

Commands to navigate to the MCP harness directory, install its dependencies, and build the project. This is a prerequisite for using the mock MCP server.

```bash
cd mcp-harness && npm install && npm run build
```

--------------------------------

### Initialize Beads and Create Tasks via Goose Prompt

Source: https://github.com/aaif-goose/goose/blob/main/documentation/docs/mcp/beads-mcp.md

This example demonstrates initializing Beads and defining tasks with dependencies using a Goose prompt. It shows the expected output for task creation and dependency management.

```goose
I want to build an expense tracker web app. Initialize beads and create all tasks with dependencies:

Frontend tasks:
- "HTML Structure" - create form, expense list, total display (no dependencies)
- "CSS Styling" - style the app (depends on "HTML Structure")
- "Wire Up UI" - connect buttons to functions (depends on "CSS Styling" AND all backend tasks)

Backend tasks:
- "Storage Setup" - localStorage helpers to save/load expenses (no dependencies)
- "Add Expense" - function to add expense to storage (depends on "Storage Setup")
- "Calculate Total" - function to sum all expenses (depends on "Storage Setup")
- "Delete Expense" - function to remove expense (depends on "Storage Setup")

Show me all tasks and which ones are ready to work on.
```

```goose
I'll initialize beads and create all the tasks with their dependencies.

─── context | beads ──────────────────────────
action: init

─── create | beads ──────────────────────────
title: HTML Structure
description: Create form, expense list, and total display for the expense tracker app
labels: frontend

─── create | beads ──────────────────────────
title: Storage Setup
description: Create localStorage helpers to save/load expenses
labels: backend

─── create | beads ──────────────────────────
title: CSS Styling
deps: ultimate-beads-test-987
description: Style the expense tracker app with CSS
labels: frontend

─── create | beads ──────────────────────────
title: Add Expense
deps: ultimate-beads-test-6gs
description: Function to add expense to storage
labels: backend

─── create | beads ──────────────────────────
title: Calculate Total
deps: ultimate-beads-test-6gs
description: Function to sum all expenses
labels: backend

─── create | beads ──────────────────────────
title: Delete Expense
deps: ultimate-beads-test-6gs
description: Function to remove expense from storage
labels: backend

─── create | beads ──────────────────────────
title: Wire Up UI
deps: ultimate-beads-test-vt2, ultimate-beads-test-5x5, ultimate-beads-test-29u, ultimate-beads-test-ae4
description: Connect buttons to functions - wire up the complete UI
labels: frontend

Here's the complete overview of your expense tracker tasks:

## 📋 All Tasks
```

--------------------------------

### Basic Goose Session

Source: https://github.com/aaif-goose/goose/blob/main/documentation/docs/guides/acp-providers.md

Starts a basic Goose session. No specific ACP provider is configured, so it may use a default or require further setup.

```bash
goose session
```

--------------------------------

### Python Virtual Environment Setup

Source: https://github.com/aaif-goose/goose/blob/main/crates/goose-mcp/src/tutorial/tutorials/first-game.md

Guides users through setting up a virtual environment for Python projects using venv. This isolates project dependencies.

```bash
python -m venv env
source env/bin/activate  # or env\Scripts\activate on Windows
```

--------------------------------

### Current Prediction Tool Call

Source: https://github.com/aaif-goose/goose/blob/main/documentation/blog/2025-08-12-mcp-testing/index.md

Example of a current prediction tool call for getting repository branches, highlighting a potential tool name mismatch.

```json
{
  "tool": "get_repo_branches", 
  "parameters": {
    "repo_owner": "block",
    "repo_name": "mcp",
    "status": "closed"
  }
}
```

--------------------------------

### Enable GitHub Extension with Token

Source: https://github.com/aaif-goose/goose/blob/main/documentation/docs/getting-started/using-extensions.md

Example of enabling the GitHub extension, which requires a personal access token. Note that this command requires Node.js and `npx` to be installed.

```bash
goose session --with-extension "GITHUB_PERSONAL_ACCESS_TOKEN=<YOUR_TOKEN> npx -y @modelcontextprotocol/server-github"
```

--------------------------------

### FileSystem Extension Example

Source: https://github.com/aaif-goose/goose/blob/main/documentation/docs/goose-architecture/extensions-design.md

A complete example of a simple FileSystem extension implementing the `read_file` tool. Demonstrates state management, asynchronous file reading, and error mapping using `ErrorData`.

```rust
use goose_macros::tool;

struct FileSystem {
    registry: ToolRegistry,
    root_path: PathBuf,
}

impl FileSystem {
    #[tool(
        name = "read_file",
        description = "Read contents of a file"
    )]
    async fn read_file(&self, path: String) -> ToolResult<Value> {
        let full_path = self.root_path.join(path);
        let content = tokio::fs::read_to_string(full_path)
            .await
            .map_err(|e| ErrorData {
                code: ErrorCode::INTERNAL_ERROR,
                message: Cow::from(e.to_string(),
                data: None,
            }))?;
            
        Ok(json!({ "content": content }))
    }
}

#[async_trait]
impl Extension for FileSystem {
    // ... implement trait methods ...
}
```

--------------------------------

### Clone and Run Goose Desktop App

Source: https://github.com/aaif-goose/goose/blob/main/ui/desktop/README.md

Steps to clone the repository, set up dependencies using Hermit, install Node.js packages, and start the development server.

```bash
git clone git@github.com:aaif-goose/goose.git
cd goose
source ./bin/activate-hermit
cd ui/desktop
pnpm install
pnpm run start
```

--------------------------------

### Set API Key and Start Session

Source: https://github.com/aaif-goose/goose/blob/main/documentation/docs/getting-started/providers.md

Demonstrates how to set the API key for a custom provider using an environment variable and then start a Goose session with that provider.

```bash
export CUSTOM_CORP_API_API_KEY="your-api-key"
goose session start --provider custom_corp_api
```

--------------------------------

### Configure Goose ACP Provider in Neovim

Source: https://github.com/aaif-goose/goose/blob/main/documentation/blog/2026-04-08-how-to-break-up-with-your-agent/index.md

Add this configuration to your Neovim setup to enable Goose as an ACP provider. Ensure Goose is installed and accessible in your PATH.

```lua
acp_providers = {
  ["goose"] = {
    command = "goose",
    args = { "acp", "--with-builtin", "developer" },
  },
},
```

--------------------------------

### Initialize Python Project with uv

Source: https://github.com/aaif-goose/goose/blob/main/crates/goose-mcp/src/tutorial/tutorials/build-mcp-extension.md

Use `uv init` to initialize a new Python project and `uv add` for package management to keep `pyproject.toml` updated.

```bash
uv init $PROJECT NAME
```

```bash
uv add
```

--------------------------------

### Install Dependencies

Source: https://github.com/aaif-goose/goose/blob/main/documentation/README.md

Installs project dependencies using npm. Run this command from the `documentation/` directory.

```bash
npm i
```

--------------------------------

### Test Docker Setup and API Accessibility

Source: https://github.com/aaif-goose/goose/blob/main/documentation/docs/tutorials/plan-feature-devcontainer-setup.md

Execute Docker Compose to start services in detached mode, list running containers, and test API endpoint accessibility.

```bash
# Test Docker setup locally
docker-compose up -d
docker ps

# Verify API accessibility
curl http://localhost:8888/api/products

# Test database connection
docker exec -it mysql mysql -u user -p api_db
```

--------------------------------

### Install Apify Local MCP Server via goose CLI

Source: https://github.com/aaif-goose/goose/blob/main/documentation/docs/mcp/apify-mcp.md

Start the Apify Local MCP server using npx. Ensure you have your Apify Token set as an environment variable.

```bash
npx -y @apify/actors-mcp-server
```

```bash
APIFY_TOKEN: <YOUR_APIFY_TOKEN>
```

--------------------------------

### Basic TypeScript MCP Server Setup

Source: https://github.com/aaif-goose/goose/blob/main/crates/goose-mcp/src/tutorial/tutorials/build-mcp-extension.md

Initialize an MCP server in TypeScript with a name and version, then connect it to a `StdioServerTransport`.

```typescript
import { McpServer } from "@modelcontextprotocol/sdk/server/mcp.js";
import { StdioServerTransport } from "@modelcontextprotocol/sdk/server/stdio.js";

const server = new McpServer({
  name: "Extension Name",
  version: "1.0.0",
});

const transport = new StdioServerTransport();
await server.connect(transport);
```

--------------------------------

### Launch Observability Server via goose Desktop

Source: https://github.com/aaif-goose/goose/blob/main/documentation/docs/mcp/cloudflare-mcp.md

Launch the Observability MCP server installer directly from the goose Desktop application. This method guides you through a series of prompts to set up the server.

```html
<a href="goose://extension?cmd=npx&arg=mcp-remote&arg=https%3A%2F%2Fobservability.mcp.cloudflare.com%2Fsse&id=cloudflare-observability&name=Cloudflare%20Observability&description=Debug%20and%20get%20insight%20into%20your%20application%27s%20logs%20and%20analytics&env=CLOUDFLARE_API_TOKEN%3DCloudflare%20API%20Token">Launch the installer</a>
```

--------------------------------

### Example Recipe Configuration (YAML)

Source: https://github.com/aaif-goose/goose/blob/main/documentation/docs/guides/recipes/recipe-reference.md

Defines a 'Code Review Assistant' recipe with parameters for language and focus, and a list of activities that utilize these parameters.

```yaml
version: "1.0.0"
title: "Code Review Assistant"
description: "Review code with customizable focus areas"
parameters:
  - key: language
    input_type: string
    requirement: required
    description: "Programming language to review"
  - key: focus
    input_type: string
    requirement: optional
    default: "best practices"
    description: "Review focus area"

activities:
  - "message: Click an option below to start reviewing {{ language }} code with a focus on {{ focus }}."
  - "Review the current file for {{ focus }}"
  - "Suggest improvements for {{ language }} code quality"
  - "Check for security vulnerabilities"
  - "Generate unit tests"
```

--------------------------------

### Generate Playwright Test Code

Source: https://github.com/aaif-goose/goose/blob/main/documentation/docs/tutorials/playwright-skill.md

This TypeScript code demonstrates how to generate a Playwright test for navigating to a specific guide within the Goose documentation. It includes setup, navigation, and assertion steps.

```typescript
import { test, expect } from '@playwright/test';

test('navigate to Using Skills guide via docs menu', async ({ page }) => {
  await page.goto('https://goose-docs.ai');
  await expect(page).toHaveTitle(/goose/);
  
  // Click on Docs in the navigation
  await page.getByRole('link', { name: 'Docs' }).click();
  
  // Expand Context Engineering category
  await page.getByRole('button', { name: 'Expand sidebar category \'Context Engineering\'' }).click();
  
  // Click on Using Skills
  await page.getByRole('link', { name: 'Using Skills' }).click();
  
  // Verify navigation
  await expect(page).toHaveURL(/using-skills/);
  await expect(page.getByRole('heading', { level: 1 })).toContainText('Using Skills');
});
```

--------------------------------

### Run Goose TUI Client (Auto-launch Server)

Source: https://github.com/aaif-goose/goose/blob/main/documentation/docs/guides/acp-clients.md

This command starts the goose TUI client and automatically launches the goose ACP server if it's installed. This is the recommended way to run the TUI.

```bash
npm start
```

--------------------------------

### Example Local .goosehints File

Source: https://github.com/aaif-goose/goose/blob/main/documentation/docs/guides/context-engineering/using-goosehints.md

Shows how to provide project-specific instructions, reference local documentation, and set workflow preferences for Goose within a specific directory.

```plaintext
This is a simple example JavaScript web application that uses the Express.js framework. View [Express documentation](https://expressjs.com/) for extended guidance.

Go through the @README.md for information on how to build and test it as needed.

Make sure to confirm all changes with me before applying.

Run tests with `npm run test` ideally after each change.
```

--------------------------------

### Basic Python MCP Server Setup

Source: https://github.com/aaif-goose/goose/blob/main/crates/goose-mcp/src/tutorial/tutorials/build-mcp-extension.md

Set up a basic MCP server in Python using `FastMCP` and `stdio_server`. The server is initialized with an extension name and then run.

```python
from mcp.server.fastmcp import FastMCP
from mcp.server.stdio import stdio_server

mcp = FastMCP("Extension Name")

if __name__ == "__main__":
    mcp.run()
```

--------------------------------

### Example Scenario Definition

Source: https://github.com/aaif-goose/goose/blob/main/evals/open-model-gym/README.md

Defines a scenario named 'file-editing' that involves creating a file and editing another. Includes setup for initial file content and validation rules to check for file existence and content matching.

```yaml
name: file-editing
description: Create and edit files
prompt: |
  1. Create joke.md containing a short joke
  2. Edit hello.rs to add a debug function

setup:
  hello.rs: |
    fn main() { println!("Hello!"); }

validate:
  - type: file_exists
    path: joke.md
  - type: file_matches
    path: hello.rs
    regex: "fn\s+debug"
```

--------------------------------

### Configure Nano Banana MCP Extension

Source: https://github.com/aaif-goose/goose/blob/main/documentation/docs/mcp/nano-banana-mcp.md

Configuration details for the Nano Banana MCP extension, including installation commands and environment variable setup for API keys. Requires Node.js for npx command execution.

```text
extensionId="nano-banana"
extensionName="Nano Banana"
description="Image generation with Gemini"
type="stdio"
command="npx"
args=["nano-banana-mcp"]
envVars=[
  { name: "GEMINI_API_KEY", label: "Gemini API Key" }
]
apiKeyLink="https://aistudio.google.com/api-keys"
apiKeyLinkText="Gemini API Key"
```

```text
name="Nano Banana"
description="Image generation with Gemini"
type="stdio"
command="npx nano-banana-mcp"
timeout={300}
envVars=[
  { key: "GEMINI_API_KEY", value: "▪▪▪▪▪▪▪▪▪▪▪▪▪▪▪▪▪▪▪▪▪▪▪▪▪▪▪▪▪▪▪▪▪" }
]
infoNote=<
  >Get your API key from <a href="https://aistudio.google.com/api-keys" target="_blank" rel="noopener noreferrer">
    Google AI Studio
  </a> and paste it in.
>
```

--------------------------------

### Whisper Installation Checklist

Source: https://github.com/aaif-goose/goose/blob/main/documentation/docs/mcp/gitmcp-mcp.md

Essential steps to install and verify the Whisper model, including Python, FFmpeg, and the Whisper package itself.

```sh
1. **✅ Install Python 3.8-3.11**
2. **✅ Install FFmpeg** for your operating system
3. **✅ Install Whisper**: `pip install -U openai-whisper`
4. **✅ Test installation**: `whisper --help`
5. **✅ Try basic transcription**: `whisper your_audio_file.mp3`
```

--------------------------------

### Start Goose Server with Network and TLS Enabled

Source: https://github.com/aaif-goose/goose/blob/main/documentation/docs/guides/environment-variables.md

Configure the Goose server to listen on all network interfaces, use a specific port, enable TLS, and set a secret key for secure communication. This is useful for remote server setups.

```bash
# Start a goosed server reachable on the local network over TLS
export GOOSE_HOST=0.0.0.0
export GOOSE_PORT=3000
export GOOSE_TLS=true
export GOOSE_SERVER__SECRET_KEY='a-long-random-secret'
goosed agent
```

--------------------------------

### Load and Start Goosed Service

Source: https://github.com/aaif-goose/goose/blob/main/documentation/docs/guides/remote-goose-server.md

These commands bootstrap the LaunchAgent and then start the `goosed` service. The `launchctl kickstart` command ensures the service is running immediately.

```bash
launchctl bootstrap gui/$(id -u) ~/Library/LaunchAgents/com.goose.goosed.external.plist
launchctl kickstart -k gui/$(id -u)/com.goose.goosed.external
```