Superagent

https://docs.superagent.sh
Admin
Superagent is an open-source framework and cloud platform for building, managing, and deploying...
Tokens:61,211
Snippets:232
Trust Score:8.4
Update:1 year ago
Show doc for...
Tokens:
Raw
### Initialize Superagent Asynchronous Python Client

Source: https://docs.superagent.sh/overview/getting-started/install-sd-ks

Example of initializing the asynchronous Superagent client in Python. This setup also requires the SUPERAGENT_API_KEY environment variable and the base URL for the Superagent API.

```Python
import os
from superagent.client import AsyncSuperagent

client = AsyncSuperagent(
    token=os.environ["SUPERAGENT_API_KEY"], # replace with your Superagent API key
    base_url="https://api.beta.superagent.sh" # or you local environment
)
```

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

### Copying Superagent Environment Configuration File

Source: https://docs.superagent.sh/overview/installation/docker-compose

This command sequence navigates to the Superagent Docker configuration directory and copies the example environment file (`.env.example`) to a new `.env` file. This new file will contain the necessary environment variables for Docker Compose to set up Superagent's core services.

```bash
cd libs/.docker
cp .env.example .env
```

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

### Copy Supabase Environment Example File

Source: https://docs.superagent.sh/overview/dependencies/supabase

Copies the `.env.example` file to `.env`, which is necessary for configuring Supabase with local environment variables before starting the services. This creates a writable configuration file based on the provided template.

```bash
cp .env.example .env
```

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

### Initialize Superagent Synchronous Python Client

Source: https://docs.superagent.sh/overview/getting-started/install-sd-ks

Example of initializing the synchronous Superagent client in Python. This requires setting the SUPERAGENT_API_KEY environment variable and specifying the base URL for the Superagent API.

```Python
import os
from superagent.client import Superagent

client = Superagent(
    token=os.environ["SUPERAGENT_API_KEY"], # replace with your Superagent API
    base_url="https://api.beta.superagent.sh" # or your local environment
)
```

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

### Complete Superagent Agent Setup and Invocation Workflow

Source: https://docs.superagent.sh/overview/getting-started/chat-with-tools

This comprehensive example combines all previous steps into a single script: initializing the Superagent client, creating an LLM, creating an agent, creating and attaching a tool, and finally invoking the agent with an input. It provides a complete end-to-end workflow for setting up and using a Superagent.sh agent.

```Python
import os
from superagent.client import Superagent

client = Superagent(
    base_url="https://api.beta.superagent.sh",
    token=os.environ["SUPERAGENT_API_KEY"]
)

llm = client.llm.create(request={
    "provider": "OPENAI",
    "apiKey": os.environ["OPENAI_API_KEY"]
})

agent = client.agent.create(
    name="Chat Assistant",
    description="An Assistant that has access to the Browser tool",
    avatar="https://mylogo.com/logo.png",  # Replace with a real image
    is_active=True,
    initial_message="Hi there! How can I help you?",
    llm_model="GPT_3_5_TURBO_16K_0613",
    prompt="Use the Browser to answer the user's question."
)

tool = client.tool.create(
    name="Browser",
    description="A portal to the internet, useful for answering questions about websites or urls.",
    type="BROWSER",
    # metadata={"key": "value"} Optional metadata for the tool.
    return_direct=False  # return_direct is required while creating a tool.
)

client.agent.add_llm(agent_id=agent.data.id, llm_id=llm.data.id)
client.agent.add_tool(agent_id=agent.data.id, tool_id=tool.data.id)

prediction = client.agent.invoke(
    agent_id=agent.data.id,
    input="Summarize superagent.sh",
    enable_streaming=False,
    session_id="my_session_id",
)

print(prediction.data.get("output"))
```

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

### Run Superagent Core Docker Services

Source: https://docs.superagent.sh/overview/installation/docker-compose

This script initializes and starts all core Superagent services using Docker Compose. It first stops any existing services, checks for and creates a Docker network if necessary, and then brings up the specified services. Users can modify the `docker-compose` command to exclude services deployed elsewhere, such as a cloud-hosted database.

```Shell
# Remove any running services
./stop.sh

# Check if the network exists
if ! docker network ls | grep -q superagent_network; then
# Create the network if it does not exist
docker network create superagent_network
fi

# Run the core services
docker-compose -f docker-compose.yml \
### Delete any services you have deployed elsewhere.
-f superagent/db/docker-compose.pgdb.yml \
-f superagent/db/docker-compose.pgadmin.yml \
-f superagent/motorhead/docker-compose.motorhead.yml \
-f ui/docker-compose.ui.yml \
up \
--build \
-d

docker logs superagent-ui
```

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

### Navigate to Weaviate Docker Directory

Source: https://docs.superagent.sh/overview/vector-stores/weaviate

Changes the current working directory to the Weaviate Docker example folder, which is necessary before running setup scripts for local Weaviate deployment.

```bash
cd libs/.docker/external/vector-store/weaviate/
```

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

### Running Superagent Core Services with Docker Compose

Source: https://docs.superagent.sh/overview/installation/docker-compose

These commands navigate to the `libs/.docker` directory and execute the `run.sh` script. This script orchestrates the deployment of Superagent's core services, including the API, UI, Motorhead, Redis, and PostgreSQL database, using Docker Compose. It provides a convenient way to start or restart all essential components.

```bash
cd libs/.docker
./run.sh
```

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

### Complete Superagent AI Agent Setup and Interaction

Source: https://docs.superagent.sh/overview/getting-started/chat-with-datasources

This comprehensive snippet combines all previous steps into a single script. It initializes the Superagent client, configures an LLM, creates an AI agent, adds a datasource, connects both the LLM and datasource to the agent, and finally invokes the agent with a query, printing the output. This serves as a full working example for setting up and using a Superagent AI assistant.

```Python
import os
from superagent.client import Superagent

client = Superagent(
    base_url="https://api.beta.superagent.sh",
    token=os.environ["SUPERAGENT_API_KEY"]
)

llm = client.llm.create(request={
    "provider": "OPENAI",
    "apiKey": os.environ["OPENAI_API_KEY"]
})

agent = client.agent.create(request={
    "name": "Earnings Assistant",
    "description": "An Assistant that's an expert at analyzing earnings reports",
    "avatar": "https://tesla.com/logo.png", # Replace with a real image
    "isActive": True,
    "llmModel": "GPT_3_5_TURBO_16K_0613",
    "initialMessage": "Hi there, how can I help you?",
    "prompt": "You are an expert at analyzing earnings reports.\nUse the earnings reports provided to answer any questions."
})

datasource = client.datasource.create(request={
    "name": "Tesla Q3 2023",
    "description": "Useful for answering questions about Tesla's Q3 2023 earnings report",
    "type": "PDF",
    "url": "https://digitalassets.tesla.com/tesla-contents/image/upload/IR/TSLA-Q3-2023-Update-3.pdf"
})

# Connect the datasource and llm to the Agent
client.agent.add_datasource(
    agent_id=agent.data.id,
    datasource_id=datasource.data.id
)
client.agent.add_llm(agent_id=agent.data.id, llm_id=llm.data.id)

# Invoke the Assistant
prediction = client.agent.invoke(
    agent_id=agent.data.id,
    input="What was Tesla's revenue?",
    enable_streaming=False,
    session_id="my_session_id"
)

print(prediction.data.get("output"))

# Tesla's revenue was 24 Billing USD according to the earnings report.
```

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

### Copy Langfuse Environment Example File

Source: https://docs.superagent.sh/overview/logging/langfuse

Duplicates the `.env.example` file to create a new `.env` file. This new file will be used to store local environment variables for the Langfuse Docker Compose setup.

```bash
cp .env.example .env
```

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

### Start Weaviate Docker Services

Source: https://docs.superagent.sh/overview/vector-stores/weaviate

Executes the provided run script to start the Weaviate services using Docker Compose. This command initiates the local Weaviate instance.

```bash
./run.sh
```

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

### cURL Example for Listing Workflow Steps

Source: https://docs.superagent.sh/api-reference/api-reference/workflow/list-steps

An example cURL command demonstrating how to make an authenticated GET request to the workflow steps API endpoint, replacing `<token>` with your actual authorization token.

```cURL
curl https://api.beta.superagent.sh/api/v1/workflows/workflow_id/steps \
-H "Authorization: Bearer <token>"
```

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

### Start Langfuse Services with Docker Compose

Source: https://docs.superagent.sh/overview/logging/langfuse

Executes the `run.sh` script, which initiates the Langfuse services and their dependencies using Docker Compose. This command brings up the entire Langfuse stack.

```bash
./run.sh
```

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

### Start Supabase Local Instance

Source: https://docs.superagent.sh/overview/installation/running-locally

Command to start the local Supabase services, including the database and other necessary components.

```Bash
supabase start
```

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

### Copy Weaviate Environment File

Source: https://docs.superagent.sh/overview/vector-stores/weaviate

Copies the example environment file (.env.example) to a new .env file. This .env file will be used by Docker Compose to configure the Weaviate service.

```bash
cp .env.example .env
```

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

### Install Superagent Python SDK

Source: https://docs.superagent.sh/overview/getting-started/install-sd-ks

Instructions for installing the Superagent Python SDK using either the poetry package manager or pip.

```Python
poetry add superagent-py
# or
pip install superagent-py
```

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

### Navigate to Supabase Docker Directory

Source: https://docs.superagent.sh/overview/dependencies/supabase

Changes the current directory to the Supabase Docker UI example folder, which is a prerequisite for running Supabase with Docker Compose. This ensures that subsequent commands are executed in the correct context.

```bash
cd libs/.docker/ui/supabase/
```

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

### Install Node.js Dependencies for Stock Assistant

Source: https://docs.superagent.sh/templates/assistants/stock-assistant

This command installs all necessary Node.js packages listed in the `package.json` file for the Stock Assistant application. It's a standard step after cloning a Node.js project to ensure all required libraries are available.

```Shell
npm i
```

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

### Start Supabase Services with run.sh Script

Source: https://docs.superagent.sh/overview/dependencies/supabase

Executes the `run.sh` script to start Supabase services using Docker Compose. This script ensures the necessary Docker network exists, stops any previously running containers, and then brings up the Supabase stack in detached mode.

```bash
./run.sh
```

```bash
# Check if the network exists
if ! docker network ls | grep -q superagent_network; then
# Create the network if it does not exist
docker network create superagent_network
fi

./stop.sh && docker compose up -d

docker logs supabase-db
```

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

### Superagent Python SDK: Agent and LLM Management Methods

Source: https://docs.superagent.sh/overview/getting-started/basic-example

This section documents the core Superagent Python SDK methods used for managing Language Models (LLMs) and AI Agents, as demonstrated in the provided example. It includes details on creating LLM configurations, creating agents, linking LLMs to agents, and invoking agents for predictions.

```APIDOC
Superagent Client Initialization:
  Superagent(base_url: str, token: str)
    - Initializes the Superagent client.
    - Parameters:
      - base_url: The base URL for the Superagent API (e.g., "https://api.beta.superagent.sh").
      - token: Your Superagent API key, typically from environment variables.

LLM Management:
  client.llm.create(request: dict)
    - Creates a new Language Model (LLM) configuration within Superagent.
    - Parameters:
      - request: A dictionary containing LLM details:
        - provider (str): The LLM provider (e.g., "OPENAI").
        - apiKey (str): The API key for the specified LLM provider.
    - Returns: An object containing the created LLM's data, including its unique 'id'.

Agent Management:
  client.agent.create(request: dict)
    - Creates a new AI Agent within Superagent.
    - Parameters:
      - request: A dictionary containing agent details:
        - name (str): The name of the agent (e.g., "Chat Assistant").
        - description (str): A brief description of the agent.
        - avatar (str, optional): A valid URL to an image for the agent's avatar (JPG or PNG).
        - isActive (bool): Whether the agent is active.
        - initialMessage (str): The first message the agent sends to a user.
        - llmModel (str): The specific LLM model to use (e.g., "GPT_3_5_TURBO_16K_0613").
        - prompt (str): The system prompt or persona for the agent.
    - Returns: An object containing the created agent's data, including its unique 'id'.

  client.agent.add_llm(agent_id: str, llm_id: str)
    - Links an existing LLM configuration to an agent.
    - Parameters:
      - agent_id (str): The unique ID of the agent.
      - llm_id (str): The unique ID of the LLM configuration.
    - Returns: (Implicitly, success/failure of the linking operation).

Agent Invocation:
  client.agent.invoke(agent_id: str, input: str, enable_streaming: bool, session_id: str)
    - Invokes the specified agent with a given input message.
    - Parameters:
      - agent_id (str): The unique ID of the agent to invoke.
      - input (str): The user's input message.
      - enable_streaming (bool): Whether to enable streaming for the response.
      - session_id (str): A unique identifier for the conversation session (best practice for user tracking).
    - Returns: A prediction object containing the agent's response, accessible via `prediction.data.get("output")`.
```

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

### Configure Superagent Vector Database with Environment Variables

Source: https://docs.superagent.sh/overview/getting-started/configuring-vector-database

This snippet demonstrates how to set environment variables to integrate a vector database like Supabase with Superagent. This method is the easiest way to get started and is suitable for single-database configurations.

```shell
export VECTORSTORE=supabase
export SUPABASE_DB_URL=postgresql://user:password@host:port/dbname
export SUPABASE_TABLE_NAME=your-table-name
```

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

### Install Project Dependencies with Poetry

Source: https://docs.superagent.sh/overview/installation/running-locally

Command to install all project dependencies using Poetry within the activated virtual environment.

```Bash
poetry install
```

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

### Start Superagent API Server with Uvicorn

Source: https://docs.superagent.sh/overview/installation/running-locally

Command to start the Superagent API server using Uvicorn with auto-reload enabled, making the API accessible locally.

```Bash
uvicorn app.main:app --reload
```

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

### Install Node.js Dependencies for Superagent UI

Source: https://docs.superagent.sh/overview/installation/running-locally

Installs all required Node.js packages and dependencies for the Superagent UI project, preparing it for development or production use.

```shell
npm i
```

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

### Start Local Supabase Services

Source: https://docs.superagent.sh/overview/installation/running-locally

Initiates the local Supabase services, including the database and authentication, which are essential for the Superagent UI to function correctly.

```shell
supabase start
```

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

### Start Superagent UI Development Server

Source: https://docs.superagent.sh/overview/installation/running-locally

Launches the Superagent UI in development mode, making it accessible locally for testing and further development.

```shell
npm run dev
```

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

### Superagent Python SDK: Create and Invoke AI Agent

Source: https://docs.superagent.sh/overview/getting-started/basic-example

This Python code snippet demonstrates the end-to-end process of setting up an AI assistant using the Superagent SDK. It covers client initialization, creating an LLM configuration, defining an AI agent with specific parameters, linking the LLM to the agent, and finally invoking the agent with a user input. API keys are loaded from environment variables.

```python
import os

from superagent.client import Superagent


client = Superagent(
    base_url="https://api.beta.superagent.sh",
    token=os.environ["SUPERAGENT_API_KEY"]
)

llm = client.llm.create(request={
    "provider": "OPENAI",
    "apiKey": os.environ["OPENAI_API_KEY"]
})

agent = client.agent.create(request={
    "name": "Chat Assistant",
    "description": "My first Assistant",
    "avatar": "https://myavatar.com/homanp.png", # A valid image URL (jpg or png)
    "isActive": True,
    "initialMessage": "Hi there! How can I help you?",
    "llmModel": "GPT_3_5_TURBO_16K_0613",
    "prompt": "You are a helpful AI Assistant",
})

client.agent.add_llm(agent_id=agent.data.id, llm_id=llm.data.id)

prediction = client.agent.invoke(
    agent_id=agent.data.id,
    input="Hi there!",
    enable_streaming=False,
    session_id="my_session" # Best practice is to create a unique session per user
)

print(prediction.data.get("output"))

# Hello there, how can I help you?
```

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

### Configure OpenAI Language Model with Superagent

Source: https://docs.superagent.sh/overview/getting-started/basic-example

This snippet demonstrates how to initialize the Superagent client and configure an OpenAI Language Model (LLM) using an API key. The configured LLM object can be reused for subsequent agents.

```python
import os

from superagent.client import Superagent


client = Superagent(
    base_url="https://api.beta.superagent.sh",
    token=os.environ["SUPERAGENT_API_KEY"]
)

llm = client.llm.create(request={
    "provider": "OPENAI",
    "apiKey": os.environ["OPENAI_API_KEY"]
})
```

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

### cURL Example to List LLMs

Source: https://docs.superagent.sh/api-reference/api-reference/llm/list

Provides a practical cURL command example to demonstrate how to make a GET request to the /api/v1/llms endpoint. This snippet includes the necessary Authorization header with a placeholder for the bearer token.

```cURL
curl https://api.beta.superagent.sh/api/v1/llms \
-H "Authorization: Bearer <token>"
```

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

### Initialize Superagent Client, Create LLM, and Agent

Source: https://docs.superagent.sh/overview/getting-started/chat-with-tools

This snippet demonstrates how to set up the Superagent client by providing the base URL and API token. It then proceeds to create an LLM object, specifying the provider and API key, and subsequently creates a new agent with a name, description, avatar, and initial message. Finally, it associates the created LLM with the agent.

```Python
import os
from superagent.client import Superagent

client = Superagent(
    base_url="https://api.beta.superagent.sh",
    token=os.environ["SUPERAGENT_API_KEY"]
)

llm = client.llm.create(request={
    "provider": "OPENAI",
    "apiKey": os.environ["OPENAI_API_KEY"]
})

agent = client.agent.create(
    name="Chat Assistant",
    description="An Assistant that has access to the Browser tool",
    avatar="https://mylogo.com/logo.png",  # Replace with a real image
    is_active=True,
    initial_message="Hi there! How can I help you?",
    llm_model="GPT_3_5_TURBO_16K_0613",
    prompt="Use the Browser to answer the user's question."
)

client.agent.add_llm(agent_id=agent.data.id, llm_id=llm.data.id)
```

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

### JSON Response Example for List Datasources

Source: https://docs.superagent.sh/api-reference/api-reference/datasource/list

An example of the JSON payload returned by the GET /api/v1/datasources endpoint upon a successful request (HTTP 200). It illustrates the structure of the data array, including nested objects for apiUser, datasources, and vectorDb.

```JSON
{
  "success": true,
  "total_pages": 1,
  "data": [
    {
      "id": "id",
      "name": "name",
      "type": "TXT",
      "apiUserId": "apiUserId",
      "createdAt": "2024-01-15T09:30:00Z",
      "updatedAt": "2024-01-15T09:30:00Z",
      "status": "IN_PROGRESS",
      "content": "content",
      "description": "description",
      "url": "url",
      "apiUser": {
        "id": "id",
        "createdAt": "2024-01-15T09:30:00Z",
        "updatedAt": "2024-01-15T09:30:00Z"
      },
      "metadata": "metadata",
      "datasources": [
        {
          "agentId": "agentId",
          "datasourceId": "datasourceId",
          "createdAt": "2024-01-15T09:30:00Z",
          "updatedAt": "2024-01-15T09:30:00Z"
        }
      ],
      "vectorDb": {
        "id": "id",
        "provider": "PINECONE",
        "createdAt": "2024-01-15T09:30:00Z",
        "updatedAt": "2024-01-15T09:30:00Z",
        "apiUserId": "apiUserId"
      },
      "vectorDbId": "vectorDbId"
    }
  ]
}
```

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

### Attach LLM to Assistant and Invoke

Source: https://docs.superagent.sh/overview/getting-started/basic-example

This snippet demonstrates how to associate a previously configured Language Model with an agent and then invoke the agent with an input message. It also shows how to retrieve and print the agent's response, using a session ID for best practice.

```python
client.agent.add_llm(agent_id=agent.data.id, llm_id=llm.data.id)

prediction = client.agent.invoke(
    agent_id=agent.data.id,
    input="Hi there!",
    enable_streaming=False,
    session_id="my_session" # Best practice is to create a unique session per user
)

print(prediction.data.get("output"))

# Hello there, how can I help you?
```

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

### Stop Langfuse Services

Source: https://docs.superagent.sh/overview/logging/langfuse

Executes the `stop.sh` script to gracefully shut down all running Langfuse services and their associated Docker containers that were started via Docker Compose.

```bash
# libs/.docker/external/observabillity/langfuse
./stop.sh
```

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

### Create a Superagent Assistant

Source: https://docs.superagent.sh/overview/getting-started/basic-example

This snippet illustrates how to create a new AI assistant within the Superagent platform. It defines various properties for the assistant, such as its name, description, type, avatar, initial message, the LLM model to use, and a custom prompt.

```python
agent = client.agent.create(
    name="Chat Assistant",
    description="My first Assistant",
    type="SUPERAGENT",
    avatar="https=//myavatar.com/homanp.png",
    is_active=True,
    initial_message="Hi there! How can I help you?",
    llm_model="GPT_3_5_TURBO_16K_0613",
    prompt="You are a helpful AI Assistant",
)
```

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

### cURL Example for Listing Datasources

Source: https://docs.superagent.sh/api-reference/api-reference/datasource/list

This cURL command demonstrates how to make a GET request to the /api/v1/datasources endpoint, including the necessary Authorization header for bearer token authentication.

```cURL
curl https://api.beta.superagent.sh/api/v1/datasources \
-H "Authorization: Bearer <token>"
```

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

### Clone Superagent Repository

Source: https://docs.superagent.sh/overview/installation/running-locally

This command sequence initiates the local setup by cloning the Superagent GitHub repository to your machine. After cloning, it navigates into the newly created 'superagent' directory, preparing the environment for further configuration.

```Shell
git clone git@github.com:homanp/superagent.git
cd superagent
```

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

### Run Stock Assistant Development Server

Source: https://docs.superagent.sh/templates/assistants/stock-assistant

This command starts the development server for the Stock Assistant application, typically using Next.js. It allows local testing and development, providing a live-reloading environment for changes.

```Shell
npm run dev
```

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

### Stop Superagent Docker Services

Source: https://docs.superagent.sh/overview/installation/docker-compose

This command stops all Docker services associated with the current project. Running this command will preserve any data linked to services that have a volume mounted, while freeing up system resources.

```Shell
./stop.sh
```

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

### Python: Create Superagent Client and Agent

Source: https://docs.superagent.sh/api-reference/api-reference/agent/create

Demonstrates how to initialize the Superagent Python client with an API token and then call the `client.agent.create` method to provision a new agent. This example includes various configuration parameters such as agent name, LLM provider, initial message, and the inclusion of a code interpreter tool.

```Python
from superagent import (
    AgentType,
    LlmProvider,
    OpenAiAssistantParameters,
    OpenAiAssistantParametersToolsItem_CodeInterpreter,
)
from superagent.client import Superagent

client = Superagent(
    token="YOUR_TOKEN",
)
client.agent.create(
    is_active=True,
    name="string",
    initial_message="string",
    prompt="string",
    llm_model="string",
    llm_provider=LlmProvider.OPENAI,
    description="string",
    avatar="string",
    type=AgentType.SUPERAGENT,
    parameters=OpenAiAssistantParameters(
        metadata={"string": {"key": "value"}},
        file_ids=["string"],
        tools=[OpenAiAssistantParametersToolsItem_CodeInterpreter()],
    ),
    metadata={"string": {"key": "value"}},
    output_schema="string",
)
```

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

### Uninstall Superagent Docker Services

Source: https://docs.superagent.sh/overview/installation/docker-compose

This command permanently removes all Docker services and their associated volumes for the current project. It's crucial to understand that this action will delete any data stored in the volumes, making it irreversible.

```Shell
./uninstall.sh
```

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

### Superagent API: List Tools Endpoint

Source: https://docs.superagent.sh/api-reference/api-reference/tool/list

Documents the GET /api/v1/tools endpoint for retrieving a list of all available tools in Superagent. It details the request parameters, authentication requirements, and the structure of the successful response, including an example JSON payload.

```APIDOC
GET https://api.beta.superagent.sh/api/v1/tools

Description: Retrieves a list of all available tools.

Headers:
  Authorization: string (Required)
    Bearer authentication of the form Bearer <token>, where token is your auth token.

Query Parameters:
  skip: integer (Optional)
    Number of items to skip for pagination.
  take: integer (Optional)
    Number of items to take for pagination.

Response (200 Retrieved):
  Content-Type: application/json
  Body:
    {
      "success": true,
      "total_pages": 1,
      "data": [
        {
          "id": "id",
          "name": "name",
          "description": "description",
          "type": "ALGOLIA",
          "returnDirect": true,
          "createdAt": "2024-01-15T09:30:00Z",
          "updatedAt": "2024-01-15T09:30:00Z",
          "apiUserId": "apiUserId",
          "metadata": "metadata",
          "apiUser": {
            "id": "id",
            "createdAt": "2024-01-15T09:30:00Z",
            "updatedAt": "2024-01-15T09:30:00Z"
          },
          "tools": [
            {
              "agentId": "agentId",
              "toolId": "toolId",
              "createdAt": "2024-01-15T09:30:00Z",
              "updatedAt": "2024-01-15T09:30:00Z"
            }
          ],
          "toolConfig": {
            "key": "value"
          }
        }
      ]
    }

Errors:
  422 Unprocessable Entity
```

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

### View Superagent Docker Service Logs

Source: https://docs.superagent.sh/overview/installation/docker-compose

These commands demonstrate how to view the real-time logs for various Superagent Docker services. The `-f` flag allows following the logs, which means new log entries will be displayed as they are generated until the command is manually stopped (e.g., with Ctrl+C).

```Shell
docker logs superagent-api -f
docker logs superagent-ui -f
docker logs motorhead -f
docker logs supabase-kong -f
docker logs langfuse-server -f
docker logs weaviate -f
```

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

### Initialize Superagent Client with API Key

Source: https://docs.superagent.sh/overview/getting-started/configuring-vector-database

This snippet demonstrates how to initialize the Superagent client using an API key and a base URL. It shows how to securely load the API key from environment variables and configure the client for either a production or local environment.

```Python
from superagent.client import Superagent

client = Superagent(
    token=os.environ["SUPERAGENT_API_KEY"], # replace with your Superagent API Key
    base_url="https://api.beta.superagent.sh" # or your local environment
)
```

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

### Initialize Superagent SDK with Local API

Source: https://docs.superagent.sh/overview/installation/running-locally

Python code snippet demonstrating how to initialize the Superagent SDK client to connect to a locally running Superagent API instance, using an API key and specifying the local base URL.

```Python
import os
from superagent.client import Superagent

client = Superagent(
    token=os.environ["SUPERAGENT_API_KEY"], # replace with your Superagent API
    base_url="http://127.0.0.1:8000" # or your local environment
)
```

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

### Configuring Local Hosts File for Docker Services

Source: https://docs.superagent.sh/overview/installation/docker-compose

This snippet illustrates how to add entries to your operating system's hosts file. These entries map local IP addresses (127.0.0.1) to the service names used within the Docker network, ensuring that services like `superagent-api` and `redis` are accessible both internally by other Docker containers and externally from your local machine via consistent URLs.

```bash
127.0.0.1  superagent-api
127.0.0.1  superagent-ui
127.0.0.1  pgdb
127.0.0.1  pgadmin
127.0.0.1  supabase-kong
127.0.0.1  langfuse-server
127.0.0.1  redis
127.0.0.1  weaviate
127.0.0.1  qdrant
```

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

### Initialize Superagent Client, Create LLM, Agent, and Tool

Source: https://docs.superagent.sh/overview/getting-started/structured-outputs

This snippet demonstrates how to set up the Superagent client using an API key, create an LLM (Large Language Model) instance with an OpenAI provider, define an AI agent with specific characteristics like name, description, and initial message, and register a browser tool. It then shows how to associate the created LLM and tool with the agent.

```Python
import os
from superagent.client import Superagent

client = Superagent(
    base_url="https://api.beta.superagent.sh",
    token=os.environ["SUPERAGENT_API_KEY"]
)

# We recommend querying for existing LLMs prior to creating.
llm = client.llm.create(request={
    "provider": "OPENAI",
    "apiKey": os.environ["OPENAI_API_KEY"]
})

agent = client.agent.create(
    name="Structured Assistant",
    description="An Assistant that returns responses in json",
    avatar="https://mylogo.com/logo.png",  # Replace with a real image
    is_active=True,
    initial_message="Hi there! How can I help you?",
    llm_model="GPT_4_1106_PREVIEW",
    prompt="Use the Browser to answer the user's question."
)

tool = client.tool.create(
    name="Browser",
    description="useful for analyzing and summarizing websites and urls.",
    type="BROWSER"
)

client.agent.add_tool(agent_id=agent.data.id, tool_id=tool.data.id)
client.agent.add_llm(agent_id=agent.data.id, llm_id=llm.data.id)
```

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

### Initialize Superagent Client and Create LLM and Agent

Source: https://docs.superagent.sh/overview/getting-started/custom-tools

This snippet demonstrates how to initialize the Superagent client using an API key, create a Language Model (LLM) object with an OpenAI API key, and then create a new agent with specific configurations like name, description, and an initial message. Finally, it shows how to associate the created LLM with the agent.

```Python
import os
from superagent.client import Superagent

client = Superagent(
    base_url="https://api.beta.superagent.sh",
    token=os.environ["SUPERAGENT_API_KEY"]
)

llm = client.llm.create(request={
    "provider": "OPENAI",
    "apiKey": os.environ["OPENAI_API_KEY"]
})

agent = client.agent.create(
    name="Stock Assistant",
    description="An Assistant that can fetch stock prices",
    avatar="https://mylogo.com/logo.png",  # Replace with a real image
    is_active=True,
    llm_model="GPT_3_5_TURBO_16K_0613",
    initial_message="Hi there, how can I help you?",
    prompt="Use the Stock API to answer the users question."
)

client.agent.add_llm(agent_id=agent.data.id, llm_id=llm.data.id)
```

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

### Get Workflow using Superagent Python Client

Source: https://docs.superagent.sh/api-reference/api-reference/workflow/get

This Python code snippet demonstrates how to retrieve a specific workflow using the Superagent client library. It initializes the client with an authentication token and then calls the 'get' method on the 'workflow' service, passing the required workflow ID.

```Python
from superagent.client import Superagent

client = Superagent(
    token="YOUR_TOKEN",
)
client.workflow.get(
    workflow_id="string",
)
```

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

### Superagent API: Get Vector Database by ID

Source: https://docs.superagent.sh/api-reference/api-reference/vector-database/get

This API documentation describes the GET endpoint for retrieving a single vector database record by its unique identifier. It outlines the required path parameters, authentication headers, and the detailed structure of the successful response body, along with potential error conditions.

```APIDOC
GET /api/v1/vector-dbs/:vector_db_id

Path Parameters:
  vector_db_id: string (Required) - The unique identifier of the vector database.

Headers:
  Authorization: string (Required) - Bearer authentication token (e.g., "Bearer <token>").

Responses:
  200 OK:
    Description: Successfully retrieved the vector database.
    Body:
      {
        "success": true,
        "data": {
          "id": "id",
          "provider": "PINECONE",
          "createdAt": "2024-01-15T09:30:00Z",
          "updatedAt": "2024-01-15T09:30:00Z",
          "apiUserId": "apiUserId",
          "options": {
            "key": "value"
          },
          "datasources": [
            {
              "id": "id",
              "name": "name",
              "type": "TXT",
              "apiUserId": "apiUserId",
              "createdAt": "2024-01-15T09:30:00Z",
              "updatedAt": "2024-01-15T09:30:00Z",
              "status": "IN_PROGRESS"
            }
          ],
          "apiUser": {
            "id": "id",
            "createdAt": "2024-01-15T09:30:00Z",
            "updatedAt": "2024-01-15T09:30:00Z",
            "token": "token",
            "email": "email",
            "agents": [
              {
                "id": "id",
                "type": "SUPERAGENT",
                "name": "name",
                "description": "description",
                "isActive": true,
                "createdAt": "2024-01-15T09:30:00Z",
                "updatedAt": "2024-01-15T09:30:00Z",
                "apiUserId": "apiUserId"
              }
            ],
            "llms": [
              {
                "id": "id",
                "provider": "OPENAI",
                "apiKey": "apiKey",
                "createdAt": "2024-01-15T09:30:00Z",
                "updatedAt": "2024-01-15T09:30:00Z",
                "apiUserId": "apiUserId"
              }
            ],
            "datasources": [
              {
                "id": "id",
                "name": "name",
                "type": "TXT",
                "apiUserId": "apiUserId",
                "createdAt": "2024-01-15T09:30:00Z",
                "updatedAt": "2024-01-15T09:30:00Z",
                "status": "IN_PROGRESS"
              }
            ],
            "tools": [
              {
                "id": "id",
                "name": "name",
                "description": "description",
                "type": "ALGOLIA",
                "returnDirect": true,
                "createdAt": "2024-01-15T09:30:00Z",
                "updatedAt": "2024-01-15T09:30:00Z",
                "apiUserId": "apiUserId"
              }
            ],
            "workflows": [
              {
                "id": "id",
                "name": "name",
                "createdAt": "2024-01-15T09:30:00Z",
                "updatedAt": "2024-01-15T09:30:00Z",
                "apiUserId": "apiUserId"
              }
            ],
            "workflowConfigs": [
              {
                "id": "id",
                "createdAt": "2024-01-15T09:30:00Z",
                "updatedAt": "2024-01-15T09:30:00Z",
                "workflowId": "workflowId"
              }
            ],
            "apiKeys": [
              {
                "id": "id",
                "name": "name",
                "displayApiKey": "displayApiKey",
                "createdAt": "2024-01-15T09:30:00Z",
                "updatedAt": "2024-01-15T09:30:00Z",
                "apiUserId": "apiUserId"
              }
            ]
          }
        }
      }
  422 Unprocessable Entity:
    Description: The request could not be processed due to semantic errors.
```

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

### Configure LLM and Create Superagent AI Agent

Source: https://docs.superagent.sh/overview/getting-started/chat-with-datasources

This snippet demonstrates how to initialize the Superagent client, create an LLM configuration, and then define an AI agent with specific parameters like name, description, and initial message. It also shows how to associate the created LLM with the agent.

```Python
import os
from superagent.client import Superagent

client = Superagent(
    base_url="https://api.beta.superagent.sh",
    token=os.environ["SUPERAGENT_API_KEY"]
)

llm = client.llm.create(request={
    "provider": "OPENAI",
    "apiKey": os.environ["OPENAI_API_KEY"]
})

agent = client.agent.create(
    name="Earnings Assistant",
    description="An Assistant that's an expert at analyzing earnings reports",
    avatar="",
    is_active=True,
    initial_message="Hi there! How can I help you?",
    llm_model="GPT_3_5_TURBO_16K_0613",
    prompt="You are an expert at analyzing earnings reports.",
)

client.agent.add_llm(agent_id=agent.data.id, llm_id=llm.data.id)
```

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

### Generate Prisma Client and Run Database Migrations

Source: https://docs.superagent.sh/overview/installation/running-locally

Commands to generate the Prisma client and apply pending database migrations to the Supabase instance, ensuring the database schema is up-to-date.

```Bash
prisma generate
prisma migrate dev
```

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

### Install Superagent Python SDK

Source: https://docs.superagent.sh/templates/assistants/chat-image-generation

This snippet demonstrates how to install the Superagent Python SDK using either Poetry or Pip. It's the first step to integrate Superagent into your Python project.

```Python
poetry add superagent-py
# or
pip install superagent-py
```

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

### Create Superagent Datasource using Python SDK

Source: https://docs.superagent.sh/overview/getting-started/configuring-vector-database

This Python code illustrates how to initialize the Superagent client and create a new datasource. It shows how to specify the datasource's name, description, type (e.g., PDF), and URL, allowing Superagent to ingest data from the specified source into the configured vector database.

```python
from superagent.client import Superagent

client = Superagent(
  token=os.environ["SUPERAGENT_API_KEY"], # replace with your Superagent API Key
  base_url="https://api.beta.superagent.sh" # or your local environment
)

datasource = client.datasource.create(request={
  "name": "Tesla Q3 2023",
  "description": "Useful for answering questions about Teslas Q3 2023 earnings report",
  "type": "PDF",
  "url": "https://digitalassets.tesla.com/tesla-contents/image/upload/IR/TSLA-Q3-2023-Update-3.pdf"
})
```

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

### Uninstall Langfuse Services

Source: https://docs.superagent.sh/overview/logging/langfuse

Executes the `uninstall.sh` script to remove all Langfuse services, containers, and associated Docker resources. This command performs a clean uninstallation of the local Langfuse setup.

```bash
# libs/.docker/external/observabillity/langfuse
./uninstall.sh
```

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

### Navigate to Langfuse Docker Compose Directory

Source: https://docs.superagent.sh/overview/logging/langfuse

Changes the current working directory to the specific location where the Langfuse Docker Compose setup files are stored within the project structure.

```bash
cd libs/.docker/external/observabillity/langfuse/
```

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

### Configure Superagent Environment Variables

Source: https://docs.superagent.sh/overview/installation/running-locally

This snippet provides an example of the mandatory and optional environment variables required to run Superagent, including API keys for OpenAI, Qdrant, AgentOps, Langfuse, LangSmith, E2B, Replicate, and database connection details for Supabase.

```Shell
OPENAI_API_KEY=<YOUR_OPENAI_API_KEY>
DATABASE_URL=postgres://<user>:<password>@<db_host_name>:<db_port>/postgres?pgbouncer=true
DATABASE_MIGRATION_URL=postgresql://<user>:<password>@<db_host_name>:<db_port>/postgres
JWT_SECRET=superagent
VECTORSTORE=qdrant
QDRANT_API_KEY=<YOUR_QDRANT_API_KEY>
QDRANT_HOST=<YOUR_QDRANT_HOST>
QDRANT_INDEX=superagent
MEMORY_API_URL=https://memory.superagent.sh
AGENTOPS_API_KEY=<YOUR_AGENTOPS_API_KEY>
AGENTOPS_ORG_KEY=<YOUR_ORG_KEY>
LANGFUSE_PUBLIC_KEY=<LANGFUSE_PUBLIC_KEY>
LANGFUSE_SECRET_KEY=<LANGFUSE_SECRET_KEY>
LANGFUSE_HOST=<LANGFUSE_HOST>
LANGCHAIN_TRACING_V2=False
LANGCHAIN_ENDPOINT=https://api.smith.langchain.com
LANGSMITH_PROJECT_ID=<LANGSMITH_PROJECT_ID>
LANGCHAIN_API_KEY=<LANGCHAIN_API_KEY>
E2B_API_KEY=<E2B_API_KEY>
REPLICATE_API_TOKEN=<REPLICATE_API_TOKEN>
```
Tokens:
Raw