Agentico v0.4.0 - Introducing MCP Agents

A new feature, Agents, with submenus for creating and managing agents, tasks, and playground (workflows for the future). This feature allows users to create custom agents that can assist them in various tasks related to building and deploying AI Agents.

Agents

• You can now create a new agent by providing a name, role, and goal. Similar to crewAI here: https://docs.crewai.com/concepts/agents
• The agent has a backstory and a list of tasks it can perform.
• The agents will be able to assist users in various tasks related to building and deploying AI Agents.

Tasks

• You can now create a new task by providing a name, description, and goal. Similar to crewAI here: https://docs.crewai.com/concepts/tasks

Playground

• The playground assists users in building, testing, and deploying agents.

Agentico v0.3.0 - Projects, Apps, Servers and Tools hierarchy

New Features

• Projects, Apps, Servers, and Tools hierarchy: The tools are now related to the projects, and generation is automatic (not manual, as in the the previous version).
• Collapsible Utility Navigation: Added collapsible functionality to the utility navigation for improved usability. (5d03f29)
• Dashboard Metrics: Dashboard now fetches metrics directly from the database. (7b3e9d0)
• Gradient Text and Icons: Enhanced UI with gradient text for headers and new icons. (2f137fe, 666ee3c)
• Project Details with Slug Support: Added slug support for project details and updated dashboard UI links. (917c8e4)
• Collapsible Actions Menu: Refactored ResourceHeader to include collapsible actions menu and code viewer toggle. (caa60a0, 17b6845)

UI/UX Improvements

• Landing Page: Improved landing page visuals and added a background. (3509139, 8c88c8c, 492ff1a)
• Dashboard UI: Enhanced dashboard layout and styling. (439b910)
• Component Refactoring: Split Dashboard page and other components into smaller, reusable parts. (cc30621, 6995ae8)

Bug Fixes

• ResourceHeader and Sidebar Menu: Fixed issues with ResourceHeader and Sidebar menu. (2a7200d)
• Tab Counters and Tree View: Corrected tab counters and tree view display issues. (204e90c)
• Import and Type Errors: Resolved import and type errors across components. (9966964)

Refactoring

• Dashboard Metrics Calculation: Improved the calculation of metric trends on the dashboard. (3880e46)

Miscellaneous

• Icons and Styling: Updated UI with new icons and gradient text for headers. (666ee3c)
• Collapsible State: Enhanced sidebar functionality with collapsible state. (255d705)

---

For a complete list of changes, refer to the commit history.

---

To Dos

• Automatically generate the Schemas and create the dependencies with Tools and Apps.
• Create the code packages to be deployed (build).
• Implement Deploy/Undeploy of servers.

0.2.0 Studio Introduction

New Features

• Studio Introduction: Welcome to the Studio! A powerful and intuitive interface for designing, managing, and deploying Agentic Workflows (WFs) at scale. 🎉
• Workflow Editor: A fully interactive editor to create and configure workflows with ease, featuring:
  • Node Picker for adding new nodes.
  • Workflow Canvas for visualizing and connecting nodes.
  • Workflow Settings Dialog for managing metadata like name and description.
• Error Handling: Enhanced error boundary with Sentry integration for monitoring and debugging workflow-related issues.

🛠 Improvements

• Save, Deploy, and Run Workflows: Added functionality to save, deploy, and execute workflows directly from the editor. Placeholders only.
• Enhanced Feedback: Real-time toast notifications for success and error states during workflow operations.
• Simulated API Calls: Improved simulation for saving, deploying, and running workflows with realistic delays.

🐛 Bug Fixes

• Resolved minor UI glitches in the workflow canvas.
• Improved error handling during workflow execution to prevent crashes.

🌟 Highlights

Agentic Workflows for MCP at Scale: Empowering users to design and deploy workflows tailored for MCP environments with unparalleled efficiency. 🤓

Enjoy the new Studio experience and take your workflows to the next level!

v0.1.0 - Project and Apps management

Orchestrate AI Agents at Scale

We're excited to announce the initial release of Agentico v0.1.0, the first official (alpha) version. This is a unified platform for deploying, managing, and optimizing AI agents across your organization.

Key features include:

MCP (Model Context Protocol)

Agentico MCP provides the tools, infrastructure, and insights you need to build robust, multi-agent workflows for your organization. The idea behind Agentico is to offer a seamless way to create agentic systems that can easily interact with each other and users. The first versions and iterations of Agentico will focus on enabling MCP to be a core component. This version includes:

• Project Management: Create and manage projects encapsulating your agentic systems.
• Application Management: Add and manage applications within your projects, allowing for modular development and deployment. Importing OAS (OpenAPI Specification) files, the user can easily create applications interacting with external APIs, enabling the MCP Clients to communicate with external services.

Demo: https://youtu.be/mhjJv-i7CrI

Deployment

The application is deployed at https://app.agentico.dev/

Installation (on-premise)

Backend services, such as Supabase, still need to be set up. For instructions on setting up your backend, please refer to the Supabase documentation.

Stay tuned for a more streamlined installation process and documentation on configuring the backend services in future releases. We plan to provide an automated setup process, including environment variable configurations and database migrations.

Enterprise support and custom deployment options are available upon request. Please feel free to contact our support team for more information.

Known Issues

You may find many. This is just the alpha version, and I appreciate all the good feedback.

Upcoming in Future Releases

We are committed to continuously improving the Agentico Dashboard. Here are some of the features and enhancements we have planned for future releases:

• Agent Management: Create and manage agents within your applications, facilitating the development of intelligent systems that can perform tasks autonomously.
• Enhanced User Interface: Introducing a workflow designer for building complex agentical systems with a visual interface. ✊🏽🤖🚀
• Import/Export Functionality: The ability to import and export projects and applications to make it easier to share and collaborate on agentic systems.
• Additional integration points with Agentico services
• Extended documentation

Feedback

We welcome your feedback! Please contact our support channels or submit issues and feature requests.

Stay tuned for more updates on this exciting adventure! 🚀✨
Go Rebels!✊🏽 🦸‍♂️

Stay tuned https://go.rebelion.la/agentico-news
Contact us https://go.rebelion.la/contact-us

---

Thank you for using Agentico Dashboard!

v0.1.0-alpha.174181

Initial PoC version with mock data. In summary, the goal is:

• Projects: Projects are the internal interfaces that allow clients to organize their AI Tools, Servers, Models, Applications, Data, and Agents. They are the main entry point for users to interact with AI Tools.
  The user can choose a project and see the AI Tools, Servers, Models, Applications, Data, and Agents associated with it.
  • The user can create a new project and associate AI Tools, Servers, Models, Applications, Data, and Agents with the project.
  • The user can manage the project's lifecycle.
• Applications: Applications are the external interfaces that allow clients to use AI tools.
  • The user can choose an application and see the AI Tools associated with the application's APIs.
  • The user can create a new application and associate AI Tools with the application's APIs.
  • The user can manage the lifecycle of the application.
• Servers: Lightweight programs that each expose specific capabilities through the standardized Model Context Protocol (MCP).
  The user can choose a server and see the AI Tools, Models, Data, and Agents associated with it.
  • The user can create a new server and associate AI Tools, Models, Data, and Agents with the server.
  • The user can manage the server's lifecycle.
• AI Tools (or Services): AI Tools are the internal interfaces that allow servers to expose executable functionality to clients and are the APIs for AI agents and clients.
  • The user can choose an AI Tool and see the Servers and Agents associated with the AI Tool.
  • The AI Tools are a 1:1 mapping to the Applications API endpoints.
  • The user can create a new AI Tool and associate Servers and Agents with the AI Tool.
  • The user can manage the lifecycle of the AI Tool.

The Applications catalog is a list of applications with APIs (OpenAPI). Users can tag applications with labels and add them to their favorites.