Infrastructure Requirements
Relevant source files
The following files were used as context for generating this wiki page:
Nexus Orchestrator is designed as a local-first, autonomous infrastructure daemon that functions as a "nervous system" for AI development. Its primary purpose is to bridge the gap between AI agents and their operating environments by managing connectivity, state, and security protocols autonomously.
The infrastructure is built upon four core pillars: connectivity via MCP, workflow orchestration through LangGraph, persistent context management, and governance using OPA and Zero Trust principles. These pillars ensure that AI agents can interact with infrastructure in a secure, stateful, and governed manner.
Sources: README.md:1-12
Core Architectural Components
The infrastructure requirement for Nexus focuses on a background orchestration layer that supports non-human identities.
Connectivity and Protocol Layer
The system requires support for the Model Context Protocol (MCP) to facilitate communication between agents and infrastructure. It specifically manages SSH connectivity to handle remote environment interactions.
Sources: README.md:4-14
Orchestration and Workflow
Workflow management is powered by LangGraph. The infrastructure must support the execution of complex graphs that define how agents interact with the environment.
Sources: README.md:15
Context and State Management
Nexus maintains a persistent state through a dedicated context layer. A key component of this is the tracking of agent identities and trust levels.
The diagram above illustrates the relationship between the Nexus Daemon and its functional layers.
Sources: README.md:3-17, context/AGENTS.md:1-10
Security and Governance Requirements
Nexus implements a Zero Trust security model specifically for non-human identities (AI agents). This requires integration with Open Policy Agent (OPA) for governance.
Agent Trust Levels
The system tracks known agents and assigns trust levels to manage permissions and access to the infrastructure.
| Component | Requirement | Description |
|---|---|---|
| Identity Tracking | Agent ID | Unique identifier for each AI agent |
| Security Policy | Trust Level | Defined trust score for OPA evaluation |
| Authorization | OPA | Policy engine for governing agent actions |
| Connectivity | SSH | Secure shell management for remote execution |
Sources: README.md:17, context/AGENTS.md:5-10
System Integration
The infrastructure functions as an "Autonomous Infrastructure Daemon," meaning it must run as a background process capable of managing its own lifecycle and the environments it oversees.
This sequence shows the interaction between an agent and the underlying infrastructure managed by Nexus.
Sources: README.md:1-10, context/AGENTS.md:1-5
Implementation Details
The project utilizes standard Node.js/JavaScript ecosystem tools for its infrastructure, as evidenced by the project structure.
- Runtime: Requires a JavaScript environment (Node.js) capable of executing the orchestration logic.
- Version Control & Ignored Files: The system manages local-first state while ignoring standard environment files and dependency folders.
- Dependency Management: Utilizes
package.jsonfor managing the lifecycle and dependencies of the orchestrator.
Sources: .gitignore:1-10, package.json
Summary
The Nexus Orchestrator infrastructure requires a robust local-first environment capable of supporting MCP and SSH protocols. It mandates a stateful context layer for tracking agent identities via AGENTS.md and a governance layer using OPA to enforce Zero Trust security for AI agents. By synthesizing LangGraph for workflows and persistent context management, Nexus provides a secure and autonomous bridge between AI logic and operational infrastructure.