# What Is DeAgentAI ## Overview of the Framework The De(cistion)Agent framework (DeAgent) is proposed to address these three core challenges. It provides a structure for creating, operating, and interacting with AI Agents within distributed systems. A DeAgent instance is defined by:

DeAgent Framework Diagram

* **Lobe:** Represents the cognitive engine of the Agent. It encapsulates the logic for invoking one or more LLMs, processing inputs, and generating outputs. Agent creators specify the Lobe(s) to be used, potentially referencing pre-defined or custom implementations. * **Memory:** Contains the Agent's initial state (genesis memory) and its accumulated interaction history (evolving memory). This provides the basis for continuity. * **Tools:** A defined set of capabilities the Agent can utilize, such as accessing external data, interacting with other systems, or performing specific actions within the distributed environment. Includes descriptions for how the Lobe can invoke them. **Creation:** A DeAgent is instantiated by publishing its definition (Lobe URI, initial Memory state, Tool specifications) to a specific address or identifier within a distributed system (e.g., a blockchain mainnet, a decentralized storage network). **Inherited Properties:** Crucially, a DeAgent naturally inherits the properties of its underlying distributed system. If the system guarantees immutability for published data, the Agent's core definition becomes immutable. If the system provides strong consensus and ordering guarantees (finality), this forms the foundation for the Agent's Identity and Continuity. **Technical Neutrality:** The DeAgent framework is agnostic to specific encoding formats, publication methods, or advanced features (like native immutability) of the underlying distributed system. However, developers and communities adopting DeAgent on a specific platform should establish clear standards and specifications. The *minimum requirement*for the host system is that it achieves a unique, final state agreed upon by participants (i.e., it possesses inherent Consensus and Identity). **Interaction:** Users interact with a DeAgent by submitting operations (e.g., transactions in a blockchain context) targeted at the Agent's address. These operations contain the interaction payload (text, image, audio, video inputs, or references to immutable data pointers if payloads exceed size limits). Platform-specific operation standards are needed. **Execution:** The DeAgent framework relies on **Executors**. These are nodes responsible for processing interactions. Executors can be centralized, run by multiple trusted parties, or fully decentralized. They monitor the distributed system for interactions targeting DeAgents they service, retrieve the Agent's definition and current state (Memory), execute the interaction using the specified Lobe and Tools, and generate potential results. **Consensus and Finalization:** Executors submit their results along with proofs of execution. A network of **Committers**(validators) then applies a consensus protocol (detailed later) to select a single canonical result for each interaction cycle, ensuring Identity. This result updates the Agent's state (Memory) on the distributed system, establishing Continuity. --- # Agent Instructions: Querying This Documentation If you need additional information that is not directly available in this page, you can query the documentation dynamically by asking a question. Perform an HTTP GET request on the current page URL with the `ask` query parameter: ``` GET https://deagentai.gitbook.io/deagentai/technical-framework/readme/publish-your-docs.md?ask= ``` The question should be specific, self-contained, and written in natural language. The response will contain a direct answer to the question and relevant excerpts and sources from the documentation. Use this mechanism when the answer is not explicitly present in the current page, you need clarification or additional context, or you want to retrieve related documentation sections.