AIMusubi — Automate real network devices with a local…

Pitch

Description

Comments

About this project

AIMusubi is a local-first automation framework that seamlessly connects LLM-driven intent-based operations to real network devices like Cisco, Arista, and VyOS. With its unified API, reproducible bootstraps, and observability features, it enables efficient network management and operations directly in your lab environment.

AIMusubi is a sophisticated local-first agentic automation framework designed to integrate LLM-driven, intent-based operations with real network devices, specifically Cisco IOS-XE, Arista EOS, and VyOS. It operates through a unified API, allowing for reproducible setups and comprehensive observability.

Key Features

Local-first Architecture: Operates entirely within a user's lab environment, ensuring that all devices, data, and APIs remain under user control.
Direct Device Integration: Seamlessly communicates with Cisco, Arista, and VyOS using their native APIs, facilitating real-world network management scenarios.
Unified Intent Engine: Executes consistent operations across different vendor platforms, simplifying multi-vendor environments.
Agentic Design: Tailored for automating workflows that utilize LLM tool-calling, enhancing the ability to leverage large language models in network management.
Reproducible Bootstraps: Simplifies environment creation through a single script, allowing users to quickly set up their frameworks.
Open-Core Structure: Provides a clear distinction between the lab framework and potential enterprise-level extensions in the future.

Full Stack Deployment

AIMusubi establishes a complete Level-5 agentic NetOps stack, which includes:

AIMusubi API (powered by FastAPI) with endpoints for executing intents, accessing metrics, and monitoring health.
Vendor Adapters for effective communication with various network devices using standard protocols like RESTCONF and JSON-RPC.
Intent Engine that supports vendor-agnostic network operations, allowing users to execute standard commands such as iface.list, routing.v4.rib, and ospf.neigh.
Observability Tools with Prometheus metrics and integration with Grafana dashboards for data visualization.
An LLM Frontend offering a web interface fully integrated with AIMusubi’s OpenAPI schema for easy interaction.

AIMusubi also bundles essential network diagnostic tools like nmap, masscan, and SNMP utilities within the lab environment, providing visibility comparable to a professional NetOps operation.

Intended Users

AIMusubi is ideal for:

Network Engineers seeking to explore LLM-driven NetOps capabilities.
Homelab Enthusiasts managing diverse network topologies across different vendors.
SRE and DevOps Engineers interested in experimenting with agentic workflows.
Educators and Students focused on mastering infrastructure automation techniques.

Project Overview

Current Version: 1.0.0 (Open-Core Lab Release)
Supported Devices: Cisco IOS-XE, Arista EOS, VyOS
Operating System Support: Ubuntu (bare-metal) and Docker environments.

For those keen to dive deeper into AIMusubi, comprehensive documentation is available, including installation guides, activation procedures, and a detailed reference on adapters and intents. Additionally, an active community can be found on Discord for support and collaboration, reinforcing the mission to innovate and transform NetOps practices.