### Quickstart (SDK, Agent-first)

Source: https://gdplabs.gitbook.io/sdk/gl-ai-agent-package/resources/reference/python-sdk

Install the SDK and get started with agent creation and execution.

```APIDOC
## Quickstart (SDK, Agent-first)

### Description
Install the `glaip-sdk` package and begin using the agent functionality.

### Installation
```bash
pip install glaip-sdk
# or for local development:
pip install glaip-sdk[local]
```

### Usage
```python
from glaip_sdk import Agent

# Create an agent instance
agent = Agent(...)

# Run the agent locally
agent.run(...)

# Optionally, deploy the agent for remote execution
agent.deploy()
```
```

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

### Quickstart Guide

Source: https://gdplabs.gitbook.io/sdk/llms-full.txt/1

A step-by-step guide to get started with the GL Connectors SDK, including installation and setting up your first integration.

```APIDOC
# Quickstart

This page assumes you have fulfilled the Prerequisites in your environment.

### Getting Started

#### Installation
```shellscript
uv add gl-connectors-sdk
```

#### Setting Up the Integration

The API Key and User Token are accessible through our console. To use the console, please see the following page: Connectors Console, or if you want to go directly to the console, it is here in Connectors Console.

In the dashboard, to activate one of the integration, we will use our Google Drive Integration to search for files.
1. Under `Google_drive` module, click `Add New Integration` button
2. Follow the generated URL and complete the authentication
3. Later, you will have an active integration for `google_drive` under your email.
```

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

### Install and Quick Start Commands

Source: https://gdplabs.gitbook.io/sdk/gl-ai-agent-package/resources/reference/cli-commands

Commands to install the SDK and perform initial setup and verification. Use `aip accounts add` for interactive credential setup and `aip status` to test connectivity.

```bash
pipx install glaip-sdk
```

```bash
aip accounts add prod   # interactive credential prompt
```

```bash
aip status              # smoke test connectivity
```

```bash
aip agents list         # browse resources (defaults to active account)
```

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

### Install and Quick Start Commands

Source: https://gdplabs.gitbook.io/sdk/llms-full.txt

Install the SDK using pipx and run initial commands to add an account, check status, and list agents. Assumes interactive credential prompts for account setup.

```bash
pipx install glaip-sdk

aip accounts add prod   # interactive credential prompt

aip status              # smoke test connectivity
aip agents list         # browse resources (defaults to active account)
```

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

### Setup Guide

Source: https://gdplabs.gitbook.io/sdk/gen-ai-sdk/resources/supported-models

Indicates that a setup guide is available for configuring the environment.

```text
Setup Guide
```

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

### Start Here: Python SDK Agent-First

Source: https://gdplabs.gitbook.io/sdk/gl-ai-agent-package/introduction-to-gl-aip

This section guides users on starting with the Python SDK using an agent-first approach. It highlights the minimal setup required for local development.

```python
glaip-sdk
```

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

### Setup UV Authentication and Install Dependencies

Source: https://gdplabs.gitbook.io/sdk/llms-full.txt/1

Executes setup scripts for UV authentication and dependency installation. Use setup.sh for Unix-based systems and setup.bat for Windows.

```bash
./setup.sh
```

```cmd
setup.bat
```

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

### Full Agent Configuration Example

Source: https://gdplabs.gitbook.io/sdk/llms-full.txt

This example demonstrates a comprehensive agent setup, including optional fields like 'version', 'description', 'metadata', 'tools', 'agents', 'mcps', 'tool_configs', and 'agent_config'.

```json
{
  "name": "data-analyst",
  "type": "config",
  "framework": "langchain",
  "version": "1.0.0",
  "description": "Analyzes data and generates reports",
  "instruction": "You are a data analyst. Analyze the provided data and generate comprehensive reports with visualizations.",
  "metadata": {
    "team": "analytics",
    "environment": "production"
  },
  "tools": ["tool-uuid-1", "tool-uuid-2"],
  "agents": ["sub-agent-uuid"],
  "mcps": ["mcp-config-uuid"],
  "tool_configs": {
    "tool-uuid-1": {
      "api_key": "analytics-api-key",
      "timeout": 30
    }
  },
  "language_model_id": "model-uuid",
  "agent_config": {
    "max_recursion_limit": 10,
    "memory": "mem0",
    "planning": true,
    "tool_output_sharing": true
  }
}
```

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

### Setup Multi-Agent Example Environment

Source: https://gdplabs.gitbook.io/sdk/gl-ai-agent-package/tutorials/multi-agent-system-patterns

Navigate to the multi-agent system patterns directory, synchronize dependencies, and set up the environment variables file.

```bash
cd gl-aip/examples/multi-agent-system-patterns
uv sync
cp .env.example .env  # then edit with your credentials
```

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

### Setup Multi-Agent Examples

Source: https://gdplabs.gitbook.io/sdk/llms-full.txt

Clone the cookbook, navigate to the multi-agent patterns directory, sync dependencies, and configure environment variables.

```bash
git clone https://github.com/gl-sdk/gen-ai-sdk-cookbook.git
cd gl-aip/examples/multi-agent-system-patterns
uv sync
cp .env.example .env  # then edit with your credentials
```

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

### Clone Example Repository

Source: https://gdplabs.gitbook.io/sdk/gl-ai-agent-package/tutorials/multi-agent-system-patterns

Clone the example repository to get started with multi-agent system patterns. This sets up the necessary project structure.

```bash
git clone https://github.com/gl-sdk/gen-ai-sdk-cookbook.git
cd gl-aip/examples/multi-agent-system-patterns

```

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

### Quickstart: Agent-first SDK Usage

Source: https://gdplabs.gitbook.io/sdk/gl-ai-agent-package/resources/reference/python-sdk

Create an Agent instance and run it locally. Optionally deploy for remote execution.

```python
from glaip_sdk.agent import Agent

agent = Agent()
agent.run()
agent.deploy()
```

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

### Install MCP Client and SDKs

Source: https://gdplabs.gitbook.io/sdk/llms-full.txt/1

Installs necessary packages for handling MCP clients and GL AI SDK locally. This setup is required for advanced MCP tool filtration examples.

```shellscript
uv init --bare
uv add glaip-sdk[local] gllm-tools-binary
```

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

### Navigate and Sync Example Directory

Source: https://gdplabs.gitbook.io/sdk/gl-aip/tutorials/multi-agent-system-patterns

Navigate to the multi-agent system patterns example directory and synchronize dependencies using uv.

```bash
cd gl-aip/examples/multi-agent-system-patterns
uv sync
```

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

### Install and Set Up WebM Recording

Source: https://gdplabs.gitbook.io/sdk/llms-full.txt/1

Install the necessary package for WebM recording and run the setup command to install Playwright Chromium. This is a one-time setup.

```bash
pip install "gl-computer-use[recording]"
gl-computer-use-setup # installs Playwright Chromium (~130 MB, one-time)
```

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

### Account Configuration Example

Source: https://gdplabs.gitbook.io/sdk/gl-ai-agent-package/tutorials/cli/commands/accounts

This example demonstrates how to configure account settings, such as API keys and endpoints, via the CLI. Ensure all required parameters are provided.

```bash
gl-ai account config --api-key "YOUR_API_KEY" --api-endpoint "https://api.example.com"
```

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

### Agent Initialization and Basic Usage

Source: https://gdplabs.gitbook.io/sdk/gl-ai-agent-package/resources/reference/python-sdk

Demonstrates how to install the SDK, import the Agent class, initialize an agent with a name and instruction, and run a simple prompt.

```python
pip install glaip-sdk

from glaip_sdk import Agent

agent = Agent(
    name="hello-agent",
    instruction="You are a helpful assistant.",
)

print(agent.run("Hello"))
```

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

### Example: Document Loader Quickstart

Source: https://gdplabs.gitbook.io/sdk/gl-ai-agent-package/guides/file-processing

Refer to the 'main_with_docproc_pdf.py' file for a quickstart and example of using document loader tools for local file processing.

```python
main_with_docproc_pdf.py
```

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

### Example Code Block

Source: https://gdplabs.gitbook.io/sdk/gl-browser-use/prerequisites

This is a generic example code block. Ensure you have the necessary environment setup before running.

```plaintext
Project: /llmstxt/gdplabs_gitbook_io_sdk_llms-full_txt
Content:
ame\":\
```

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

### Running an Agent - Example

Source: https://gdplabs.gitbook.io/sdk/gl-aip/guides/agents-guide

This example demonstrates how to run an agent. Ensure the agent is properly configured before execution.

```javascript
)\"\n166:[\
```

```javascript
)\"\n167:[\
```

```javascript
)\"\n168:[\
```

```javascript
)\"\n169:[\
```

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

### Get Search Results

Source: https://gdplabs.gitbook.io/sdk/gl-smart-search/guides/sdk/web-search

Perform a search query using the SDK. This example shows how to get results for a given query.

```javascript
const results = await search(\"your search query\");
console.log(results);
```

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

### Run Privacy Quickstart Script

Source: https://gdplabs.gitbook.io/sdk/gen-ai-sdk/tutorials/security-and-privacy/pii-masking

Execute the privacy quickstart Python script. This is a common way to initiate the PII masking process using the installed library.

```bash
python privacy_quickstart.py
```

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

### Install SDK with All Extras

Source: https://gdplabs.gitbook.io/sdk/llms-full.txt/1

Install the SDK with all available extras for the full feature set.

```bash
pip install "gl-computer-use[all]"
```

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

### Get TTS Count API Example (Bash)

Source: https://gdplabs.gitbook.io/sdk/gl-speech/rest-api-reference/text-to-speech

This bash command uses curl to make a GET request to the TTS count endpoint, filtering results by date.

```bash
curl -X GET 'https://tts-api.stg.prosa.ai/v2/speech/tts/count?from_date=2024-01-01' \
```

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

### Install gl-computer-use[observability]

Source: https://gdplabs.gitbook.io/sdk/gl-computer-use/prerequisites

Use this command to install the GL Computer Use SDK with observability features. Ensure pip is up-to-date.

```bash
pip install gl-computer-use[observability]"
```

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

### Skill Instructional Content Example

Source: https://gdplabs.gitbook.io/sdk/gl-connectors/sdk/connectors-skills/creating-a-skill

Provides a template for the core instructions and an example of input/output for a custom skill. This section guides the skill's behavior and demonstrates its functionality.

```markdown
# My Skill

[Your instructions here — what to do, how to format it, what rules to follow]

## Example

Input: [example input]
Output: [example output]
```

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

### Run Desktop Automation Example

Source: https://gdplabs.gitbook.io/sdk/gl-computer-use/getting-started

This snippet demonstrates how to initiate desktop automation using the SDK. Ensure necessary configurations are in place before execution.

```javascript
console.log("Run Desktop Automation")
```

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

### Audio Input and Output Configuration Example

Source: https://gdplabs.gitbook.io/sdk/gl-ai-agent-package/guides/audio-interface

A comprehensive example demonstrating the configuration of both audio input and output, including enabling, setting sample rates, and adding event listeners.

```javascript
const agent = new Agent();

// Configure input
agent.audio.input.enable();
agent.audio.input.setSampleRate(16000);
agent.audio.input.addEventListener('data', (data) => {
  console.log('Received audio data:', data);
});
agent.audio.input.addEventListener('error', (error) => {
  console.error('Input error:', error);
});

// Configure output
agent.audio.output.enable();
agent.audio.output.setSampleRate(16000);
agent.audio.output.addEventListener('playback-end', () => {
  console.log('Playback finished.');
});
agent.audio.output.addEventListener('error', (error) => {
  console.error('Output error:', error);
});

console.log('Audio input enabled:', agent.audio.input.isEnabled());
console.log('Audio output enabled:', agent.audio.output.isEnabled());
```

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

### Start Google Realtime Session

Source: https://gdplabs.gitbook.io/sdk/gen-ai-sdk/tutorials/inference/realtime-session

Initializes and starts a realtime inference session with a specified Gemini model. Ensure you have the dotenv library installed and your environment variables configured.

```python
from dotenv import load_dotenv
load_dotenv()

import asyncio
from gllm_inference.realtime_session import GoogleRealtimeSession

realtime_session = GoogleRealtimeSession(model_name="gemini-2.5-flash-native-audio-preview-12-2025")
asyncio.run(realtime_session.start())
```

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

### Speech to Text API Request Example

Source: https://gdplabs.gitbook.io/sdk/gl-speech/rest-api-reference/speech-to-text

Example of how to make a GET request to the Speech to Text API using curl. Ensure you replace 'your-api-key' with your actual API key.

```bash
curl -X GET 'https://asr-api.stg.prosa.ai/v2/speech/stt/2fec34e1-efb1-46f7-a743-1cb35b64550d' \
 -H 'x-api-key: your-api-key'
```

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

### Global Configuration Example

Source: https://gdplabs.gitbook.io/sdk/llms-full.txt

An example demonstrating the structure of global configurations, including tool settings.

```json
{
  "tool_configs": {
    "tool-uuid-1": {
      "mode": "dry_run"
    }
  }
}
```

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

### Install GDP Labs SDK

Source: https://gdplabs.gitbook.io/sdk/gen-ai-sdk/tutorials/inference/realtime-session

Install the GDP Labs SDK using pip with the provided extra index URL. This command ensures you get the correct version for your project.

```bash
IN ( 'gcloud auth print-access-token' ) DO pip install --extra-index-url "https://oauth2accesstoken:%T@glsdk.gdplabs.id/gen-ai-internal/simple/"
```

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

### Full Connector SDK Example

Source: https://gdplabs.gitbook.io/sdk/gl-smart-search/guides/sdk/connector-search

Demonstrates a complete workflow including authentication, searching, and connecting to a connector using the SDK. Ensure environment variables for SMARTSEARCH_BASE_URL and SMARTSEARCH_TOKEN are set.

```python
import asyncio
import json
import os
from dotenv import load_dotenv

from smart_search_sdk.connector.client import ConnectorClient
from smart_search_sdk.connector.models import (
    ConnectorRequest,
    ConnectorConnectRequest,
    AppName,
)

load_dotenv()

async def main():
    client = ConnectorClient(base_url=os.getenv("SMARTSEARCH_BASE_URL"))
    await client.authenticate(token=os.getenv("SMARTSEARCH_TOKEN"))
    gl_token = os.getenv("GL_CONNECTORS_USER_TOKEN", "")

    # Search
    request = ConnectorRequest(query="List all my upcoming meetings today")
    result = await client.search_connector(
        app_name=AppName.GOOGLE_CALENDAR,
        gl_token=gl_token,
        request=request,
    )
    print("=== Connector Search ===")
    print(json.dumps(result, indent=4))

    # Connect
    connect_request = ConnectorConnectRequest(

```

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

### Include Example Code Configuration

Source: https://gdplabs.gitbook.io/sdk/gl-ai-agent-package/guides/programmatic-tool-calling

Use the 'include_example' parameter to determine whether example code should be included in the prompt. This is a boolean setting.

```text
include_example: Whether to include example code in the prompt
```

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

### Complete HITL Example with Smart Approver

Source: https://gdplabs.gitbook.io/sdk/llms-full.txt

A comprehensive example demonstrating the integration of a smart approver function with the RemoteHITLHandler. This setup includes necessary imports and defines the approval logic.

```python
import os
from glaip_sdk import Client
from glaip_sdk.hitl.remote import RemoteHITLHandler
from glaip_sdk.hitl.base import HITLRequest, HITLResponse, HITLDecision

# Approval callback
def smart_approver(request: HITLRequest) -> HITLResponse:
    # Safe tools
    if request.tool_name in ["read_file", "search"]:
        return HITLResponse(decision=HITLDecision.APPROVED)

    # Dangerous tools
    if "delete" in request.tool_name.lower():
        return HITLResponse(
            decision=HITLDecision.REJECTED,
            operator_input="Dangerous operation blocked"
        )

    # Default
    return HITLResponse(decision=HITLDecision.APPROVED)
```

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

### Run the Interactive Setup for Connector MCP Cookbook

Source: https://gdplabs.gitbook.io/sdk/gl-connectors/sdk/agentic-tools-and-model-context-protocol-mcp/connector-mcp-cookbook

Execute the interactive Python script to set up the Connector MCP Cookbook. This will guide you through the setup process, including authentication and authorization.

```bash
uv run example_interactive.py
```

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

### Agent Initialization Example

Source: https://gdplabs.gitbook.io/sdk/gl-ai-agent-package/guides/topics

Demonstrates how to initialize an Agent using the SDK. This is the default path for development.

```javascript
Agent(...)
```

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

### Install gl-computer-use[all]

Source: https://gdplabs.gitbook.io/sdk/gl-computer-use/prerequisites

Use this command to install the GL Computer Use SDK with all available features. Ensure pip is up-to-date.

```bash
pip install gl-computer-use[all]"
```

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

### Install SDK with Pip

Source: https://gdplabs.gitbook.io/sdk/tutorials/inference/realtime-session

Use pip to install the SDK, specifying an extra index URL for private packages. Ensure your environment is set up correctly, for example, using a Conda environment.

```bash
pip install --extra-index-url https://oauth2accesstoken$$gcloud$$
```

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

### Install GL Computer Use SDK

Source: https://gdplabs.gitbook.io/sdk/gl-computer-use/prerequisites

Install the core GL Computer Use package and optional extras for specific functionalities.

```bash
pip install gl-computer-use
```

```bash
pip install "gl-computer-use[recording]"
pip install "gl-computer-use[agents]"
pip install "gl-computer-use[opensandbox]"
pip install "gl-computer-use[minio]"
pip install "gl-computer-use[observability]"
pip install "gl-computer-use[all]"
```

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

### Configure PGVector Client

Source: https://gdplabs.gitbook.io/sdk/gen-ai-sdk/resources/supported-vector-data-store

Example of configuring the PGVector client. This setup uses a persistent client.

```javascript
import pg from "pg";

const pool = new pg.Pool({
  user: "your_user",
  host: "your_host",
  database: "your_database",
  password: "your_password",
  port: 5432,
});

async function embedding() {
  const result = await pool.query(
    "SELECT * FROM documents ORDER BY embedding <-> $1 LIMIT $2",
    [
      "[0.1, 0.2, 0.3, 0.4, 0.5, 0.6, 0.7, 0.8, 0.9, 1.0]::vector",
      10,
    ]
  );
  console.log(result.rows);
}

embedding();
```

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

### Run Setup Script

Source: https://gdplabs.gitbook.io/sdk/llms-full.txt/1

Executes the setup script to prepare the project environment. Use './setup.sh' for Unix-based systems (Linux, macOS) and 'setup.bat' for Windows.

```bash
./setup.sh
```

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

### Configure Qdrant Client

Source: https://gdplabs.gitbook.io/sdk/gen-ai-sdk/resources/supported-vector-data-store

Example of configuring the Qdrant client. This setup uses a persistent client.

```javascript
import { QdrantClient } from "@qdrant/js-client-rest";

const client = new QdrantClient({
  url: "YOUR_QDRANT_URL",
  apiKey: "YOUR_QDRANT_API_KEY",
});

async function embedding() {
  const result = await client.search("my_collection", {
    vector: [0.1, 0.2, 0.3, 0.4, 0.5, 0.6, 0.7, 0.8, 0.9, 1.0],
    limit: 10,
  });
  console.log(result);
}

embedding();
```

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

### Quickstart: OpenAI Text Embedding Response

Source: https://gdplabs.gitbook.io/sdk/gen-ai-sdk/tutorials/inference/em-invoker

Example output of the vectorized text from the OpenAIEM.TEXT_EMBEDDING_3_SMALL model.

```text
Vectorized text:
[-0.010044993832707405, ..., 0.0008305096416734159]
```

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

### Make a GET request to the Speech to Text API

Source: https://gdplabs.gitbook.io/sdk/gl-speech/rest-api-reference/speech-to-text

This example demonstrates how to make a GET request to the Speech to Text API endpoint using curl. Ensure you have the correct URL and any required headers.

```bash
curl 'https://asr-api.stg.prosa.ai/v2/speech/stt/models'
```

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

### Bash Command to Run Example

Source: https://gdplabs.gitbook.io/sdk/llms-full.txt

Executes the router example script using uv unicorn. Ensure environment variables are set.

```bash
uv run router/main.py
```