Devices & Inventory Dashboard
Relevant source files
The following files were used as context for generating this wiki page:
Overview
The Devices & Inventory Dashboard provides a centralized, live view of your entire network infrastructure, acting as the "single source of truth" for hardware and logical assets. It solves the problem of fragmented visibility by aggregating data from multi-vendor environments—including routers, switches, firewalls, and cloud platforms—into a unified console.
The platform is designed with a local-first and self-hosted philosophy, ensuring that your sensitive network data never leaves your infrastructure. It supports a wide range of vendors and protocols through a specialized plugin system. Every action follows a strict security model: the AI can discover and analyze your network, but a human must approve every change, and all AI reasoning is visible and audited.
Supported Vendors and Protocols
The platform connects to and manages infrastructure across:
- Networking: Cisco (IOS, IOS-XE, NX-OS), Juniper JunOS, Arista EOS.
- Security: Palo Alto PAN-OS, Fortinet FortiOS.
- Load Balancing & DDI: F5 BIG-IP, BlueCat, Infoblox.
- Cloud & Virtualization: AWS, Azure, VMware.
- Discovery Methods: SNMP, SSH, LLDP, CDP, and vendor APIs.
Sources: README.md:3-8, CLAUDE.md:14-35, backend/scripts/seed_smb_scenario.py:44-48
Specialized AI Team Members (Agents)
The platform uses "Agents," which are best understood as specialized team members you can talk to in plain English. For inventory and dashboard tasks, the primary agents involved are:
- Discovery Agent: The team member that scans your network (via SNMP/SSH) to find new devices and map their connections.
- Documentation Agent: Responsible for generating inventories, topology diagrams, and runbooks from live data.
- Master Architect Agent: The supervisor who listens to your request and routes it to the correct specialist.
Sources: CLAUDE.md:46-56, backend/app/agents/documentation/agent.py:42-53
Architecture
The platform consists of seven core components that work together to manage your inventory and provide the dashboard experience:
| Component | Role | Plain English Description |
|---|---|---|
| frontend | User Interface | The "Ops Console" where engineers chat with the AI and view inventory tables. |
| api | Traffic Controller | Handles user logins and coordinates requests between the UI and the background engines. |
| worker | Background Engine | Runs discovery scans, takes configuration backups, and handles long-running tasks. |
| postgres | Database | The main system of record that stores device details, credentials, and audit logs. |
| neo4j | Topology Map | A "living topology map" that tracks relationships between devices, subnets, and VLANs. |
| redis | Task Manager | Helps the API and workers communicate by managing the queue of pending jobs. |
| ollama | AI Brain | Runs the local AI model (LLM) so no network data is sent to external AI providers. |
Sources: README.md:32-40, frontend/README.md:3-6, deploy/docker/README.md:7-13
Vendor Plugin System
The platform uses a standardized "connector" (plugin) for each vendor. This ensures that whether the platform is talking to a Cisco router or an Arista switch, it translates the vendor-specific command output into a consistent, normalized format. This normalization allows the dashboard to show uniform data across different hardware types.
Sources: README.md:42-45, backend/app/plugins/vendors/cisco_nxos/parsers.py:1-15
Deployment Options
- Docker Compose: Ideal for lab environments or small teams. It runs the entire platform on a single server using a single command.
- Kubernetes/Helm: Designed for production use. This provides high availability across multiple nodes, ensuring the platform stays up even if a single server fails.
Sources: README.md:52-59, deploy/kubernetes/netops/README.md:9-17
Inventory Data Model
The platform maintains a comprehensive "knowledge graph" of your network. Unlike a simple spreadsheet, it understands how different components relate to one another.
How the Platform Sees Your Network
This diagram shows the relationships the platform tracks to build a complete inventory.
Sources: CLAUDE.md:67-73, backend/scripts/seed_smb_scenario.py:151-200, backend/tests/agents/documentation/test_documentation_agent.py:48-115
Change Approval Workflow
Every change requested through the dashboard follows a strict "Human-in-the-Loop" workflow to prevent accidental outages.
The full lifecycle from a plain-English request to execution.
Sources: CLAUDE.md:12, README.md:8, backend/app/llm/prompts/init.py:187-205
System Overview & Agent Routing
When you interact with the dashboard, your query is analyzed and routed to a specialist team member.
System Connectivity
Agent Routing Flow
How the Master Architect selects the correct specialist based on your goal.
Sources: CLAUDE.md:46-56, backend/app/llm/prompts/init.py:220-310, backend/app/agents/documentation/agent.py:65-79
Conclusion
The Devices & Inventory Dashboard transforms raw network data into an interactive, AI-powered operations center. By combining multi-vendor support, a secure local-first architecture, and a human-approved change workflow, it empowers network teams to manage complex infrastructure with greater visibility and safety. Through the coordination of specialized Agents and a detailed knowledge graph, the platform ensures that every device, interface, and route is accounted for and easily accessible.