AgentActs.eth

AI agent builders today are too restrictive compared to the open canvas they should have been. Each organisation is trying to eat the most by creating centralised platforms for the same.

Although the markets and developers have clearly rejected such an approach, AI agents should be launched on a platform that is decentralised and where ownership of these agents lies in the hands of the creator and the creator only. This is what we want to work on.
Just like anyone can do a lookup of a website in their browser, we want to build an index where people own AI agents that others can access by paying to use them (pay-per-use).

We have used MCPs (Model Context Protocol) to launch AI agents. These AI agents have access to many of the Model Context Protocols, allowing them a higher range of functionality. There are already 3000+ AI agents existing, allowing people to automate any workflow they want to.

AI agents have reached a peak where they can do a lot of tasks, and we want to ensure that the yield generated from such efficiency is not just in the hands of a few big companies but small developers as well. This will be done by democratising access and decentralising power.

At the core of the platform is the integration of multiple multi-context protocols (MCPs), which enable seamless interoperability across various tools and data sources. By combining these MCPs, the platform allows users to create workflows that can interact with each other. MCPs are by default open source.

In addition to MCPs, we've integrated key partner technologies like Polygon, 1inch, and ENS, each bringing their own valuable capabilities:

Our Partners for this Hackathon, Polygon serves as the robust blockchain infrastructure that powers the platform, providing scalability and low transaction costs. The integration with Polygon ensures that AI agents are deployed on-chain, enabling true ownership and transparency, and opening up a wide range of use cases in decentralized finance and digital asset management.

1inch enhances the platform’s DeFi capabilities, particularly in enabling cross-chain swaps. Through 1inch’s Fusion+ and portfolio tracking APIs, agents can facilitate token swaps across different blockchains and offer real-time insights into portfolio performance. This allows users to automate complex DeFi tasks and optimize their investments without manual intervention.

ENS (Ethereum Name Service) adds an essential layer of decentralized identity and verifiable ownership to the platform. By integrating ENS, users can assign Web3 identities to their agents, making them easily recognizable and accessible on-chain, while also fostering greater trust and usability in decentralized applications.

Python API (Dockerized): The backend API, developed in Python, serves as the communication bridge between users and the AI agents. This API handles all interactions, processing user prompts and coordinating tasks across various MCPs and external services. To ensure scalability, we have dockerized the Python API, enabling easy deployment and maintenance in any environment. Dockerizing the API also ensures that it can be run seamlessly across multiple servers, making the platform highly available and resilient.