A forma MAIS FÁCIL de configurar o ClawdBot (MoltBot) de forma “SEGURA” no PC (Docker)

How to Install Cloud Bot for Free

Introduction to Cloud Bot

• The video introduces a secure and simple method to install Cloud Bot for free, emphasizing its value as an agent for 2026 that operates continuously.

• Acknowledges that not everyone has access to a VPS or old computer, hence the tutorial will focus on installing it directly on personal computers using Docker.

Transition from Cloud Bot to Mbot

• The name change from Cloud Bot to Mbot is explained due to trademark issues with Antropic, the creator of the original model.

• Clarifies that while the name may vary (Cloudbot vs. Mbot), both refer to the same application, which is one of the best AI programming models available.

Understanding Docker

• Introduces Docker as a tool that allows users to create isolated environments (containers) for applications, simplifying installation and management.

• Discusses how Docker can streamline processes by allowing multiple containers (e.g., N8N and databases) without interference between them.

Benefits of Using Containers

• Each container operates independently with its own configuration files, making it easier to replicate environments without affecting the main system.

• Emphasizes security benefits; containers can be configured not to access each other's data or interfere with the host system.

Installation Steps Overview

• Outlines that Docker will be installed on Windows, creating an isolated environment for running Cloud Bot safely.

• Provides instructions on downloading Docker Desktop from Docker.com and selecting the appropriate version based on system specifications.

Installing Docker

• Walkthrough of installation steps including accepting terms of use and setting up Windows Subsystem for Linux (WSL).

• Notes potential issues during installation but reassures viewers about troubleshooting steps taken in real-time.

Creating Your First Container

• After successful installation, viewers are guided through downloading a "Docker File," which serves as a recipe for building their container.

• Explains what goes into this file: using Ubuntu 22.04 as a base image and pre-installing necessary packages like Curl and Git.

Setting Up the Cloudbot Environment

Initial Setup and Configuration

• The speaker discusses installing Vin, a text editor necessary for upcoming tasks, while configuring the default directory of the container.

• A folder containing a single Docker file is opened in the terminal (PowerShell), emphasizing that installation steps are straightforward and previously detailed in another video.

• The command Docker build -t cloudbot . is executed to create an image from the Docker file, indicating successful completion by checking Docker images.

Running the Container

• The speaker runs the container using Docker Run, highlighting specific port configurations and naming conventions before executing the command.

• Upon running, confirmation is given that the container has started successfully; access to its terminal is achieved through Docker's interface.

Exploring Container Contents

• Inside the terminal, a Linux environment structure is displayed using ls -l, confirming entry into a mini-system prepared for Cloudbot operations.

• Instructions are provided to install Cloudbot (now Modbot), with emphasis on selecting appropriate commands based on operating systems (Windows vs. Mac/Linux).

Installing and Configuring Cloudbot

Installation Process

• The installation of Mebot begins after copying and pasting a command from Cloudbot’s website; quick onboarding options are discussed.

• Users are prompted to select models for their agent's brain; recommendations include using Open AI's Plus plan or opting for free alternatives like Open Houter.

API Key Configuration

• Instructions follow for obtaining an API key from Open Houter, including account creation steps via Google or email sign-up.

• After creating an API key named "cloud bot," users must input this key into their configuration settings.

Model Selection and Application Setup

• Discussion about potential issues with free models suggests exploring alternative solutions if certain models become unavailable during use.

• Users can configure input applications such as Telegram or WhatsApp; however, details on this process refer back to previous videos for clarity.

Finalizing Configuration Steps

Completing Setup

• Final setup involves entering model names manually if not found automatically; users confirm selections through prompts in the terminal.

• The speaker mentions configuring skills and package managers like NPM as part of completing Cloudbot’s setup process.

Configuring Skills and APIs

Initial Setup of Skills

• The speaker discusses the dependencies for certain skills, indicating that they will skip some configurations for now.

• Specific API keys are mentioned: Google Places API, Gemini API for image generation, OpenAI's image and audio generation APIs, and Eleven Labs' audio generation API—all set to "no" for this setup.

Enabling Hooks

• The speaker enables three options related to hooks, referencing a previous video for detailed explanations.

• A token is generated during the onboarding process that needs to be saved for future use.

Internet Configuration

• Instructions are provided on how to configure internet search capabilities using the Brave browser API; by default, it does not search online.

Running the Gateway

Executing Commands

• To run the gateway, a command is executed in the terminal: cloudbot gateway --port <address>.

• All commands used in this process will be available in a file linked in the description.

Interacting with the Agent

• The speaker explains how to interact with the agent via different interfaces: Telegram, TUI (Terminal User Interface), and web interface.

Using Terminal User Interface (TUI)

Accessing TUI

• A new PowerShell tab is opened to access the container using Docker commands.

• The command cloud t UI is executed to start interacting with the agent through TUI.

Initial Interaction

• Upon greeting the agent with "Oi," it responds affirmatively but slowly due to being on a free plan.

• The speaker notes zero credits available on Open Houter but confirms ongoing interaction without cost.

Accessing Web Interface

Transitioning from TUI

• To switch from TUI back to web interface, an existing session can be terminated using Ctrl+C commands.

Opening Dashboard

• A command (cloudbot dashboard) is issued to open up a user interface dashboard along with its token.

Addressing Connection Issues

• An error occurs when trying to access via URL due to container isolation; local host cannot directly connect without proper configuration.

Configuring Container Settings

Editing Configuration Files

• Inside PowerShell, hidden files are listed; navigation leads into a directory containing configuration files relevant for cloudbot settings.

Using Text Editor within Container

• The speaker uses vim as a text editor within Docker since no graphical interface exists.

Configuring Cloudbot Gateway

Modifying Network Configuration

• The speaker discusses changing the loopback binding in the network configuration to allow the service to accept connections from the computer. This is a basic method of making it functional, although there are more advanced configurations available.

Editing Files in Terminal

• Instructions are provided on how to enter insert mode in a text editor using 'I' and navigate through the file with keyboard commands instead of a mouse. The speaker replaces "loopback" with "l".

Saving Changes and Restarting Services

• After editing, the speaker exits insert mode and saves changes by typing ':wq'. They then restart the Cloudbot service by terminating it with Ctrl+C and running it again.

Accessing Web Interface

• The speaker explains how to access the web interface at 127.0.0.1 but needs to change the port number to 18790 for proper redirection within the container.

Authorizing Devices

• An error message indicates that authorization is needed for device connection. The speaker shows how to approve devices pending authorization using specific command syntax.

Using Docker Effectively

Learning Through Interaction

• The speaker emphasizes learning about functioning as an assistant through interaction, suggesting users ask questions about preferences like names or emojis.

Security Considerations

• A warning is given regarding security; users should avoid mounting local drives or USB devices for safety reasons, highlighting that security requires proactive measures rather than simple solutions.

Starting Docker Services Easily

• To simplify starting services, users can use Docker's graphical interface instead of command line inputs, making it more user-friendly.

Utilizing Free Credits on Google Cloud

• Users are advised on creating a new Google account to take advantage of free credits offered by Google Cloud, which can help mitigate costs while experimenting with services.

Setting Up Free Trials Safely

• Guidance is provided on signing up for free trials without incurring charges by using temporary credit cards and ensuring two-step authentication is enabled for added security during setup.

How to Use Google AI Studio and Gemini API

Overview of Google AI Studio Options

• The speaker introduces the use of the Gemini API within Google AI Studio, highlighting three model options: Gemini Piki, Antigravity, and Diminis Cali.

• The discussion includes authentication methods (Oalf) for linking the API to existing plans in Docker containers, noting potential configuration issues.

Setting Up Your Project

• To utilize the Gemini API Key, users must navigate to Google AI Studio and create an API key by selecting a project. If no projects exist, they need to create one first.

• Users are instructed on how to name their project and ensure payment is made prior to creating it in Google Cloud Console.

Generating and Using the API Key

• After creating a project, users can generate an API key which can be copied for later use. The speaker emphasizes that this step is crucial for accessing models.

• Once the key is copied, users select their desired model (e.g., 2.5 flash), with a reminder that different models have varying costs associated with usage.

Understanding Pricing Structure

• In Google Studio's Playground section, users can filter models like Gemini 2.5 Flash and view pricing details—30 cents per million tokens for input and $2.50 for output.

• The speaker expresses enthusiasm about potentially being the first to demonstrate running this setup in Docker while encouraging viewers to engage with the content through likes and comments.

Conclusion

• The video aims to provide valuable insights into using Google's tools effectively while fostering community interaction through viewer engagement prompts.