Integration & Extensibility

MCP Bidirectional Integration

Open NetBrain’s network intelligence to the entire AI ecosystem, and bring external AI into your network. Model Context Protocol (MCP) support transforms NetBrain into a first-class citizen of the enterprise AI stack. 

As an MCP Server, NetBrain exposes its unique network topology, device state, and diagnostic context to any external AI tool, from Claude to custom LLMs. As an MCP Client, NetBrain pulls in signals from APM platforms, observability tools, and ITSM systems, giving its own agents richer, cross-domain context for every investigation. 

NetBrain as MCP Server  (outbound to external AI) 

Exposes network topology, device state, configurations, intents, and diagnostic context to any external AI tool that speaks MCP. Claude, GPT-4, Gemini, or any custom LLM connects with no custom integration work. Your AI ecosystem inherits real-time network context on day one. 

• Real-time multi-vendor topology, refreshed continuously across hybrid, multi-cloud, and SD-WAN. 
• Device configurations, Golden Configs, and detected drift. 
• The Network Intent library, holding your network’s design as enforceable rules. 
• Live forwarding paths and the historical baseline to compare against. 
• The One-IP Table: every IP, MAC, and switch port traced to its true L2/L3 gateway. 

Access is read-by-default. Any write operation routes through NetBrain’s existing approval workflows, the same guardrails that govern Deep Diagnosis and Change Agent. 

NetBrain as MCP Client  (inbound from APM, observability, ITSM) 

Pulls signal from adjacent systems directly into NetBrain agents. The network agent doesn’t have to guess what’s happening above the network: it sees the APM trace, the observability metric, the ticket context, all as part of one investigation. 

• APM platforms feed application performance data into Deep Diagnosis. 
• Observability tools (Splunk, ThousandEyes, BMC) feed metric signals. 
• ITSM systems (ServiceNow, Jira) feed ticket history and ownership context. 
• Cross-domain RCA in a single pass: the network agent reasons about both network state and upstream signals at once. 

Model Flexibility

Run NetBrain AI on any foundation model, cloud, private, or air-gapped. Claude, Gemini, GPT-4, and any OpenAI-compatible API endpoint, including self-hosted and fully air-gapped deployments. The answer to “AI vendor lock-in” in every sales conversation: there isn’t any. Switch models without reconfiguring agents, rebuilding Skills, or changing integrations. 

API Authenticator

Centralize credentials for every network API: OAuth 2.0, API Key, Basic Auth. API Authenticator provides a unified, secure credential vault for all external API integrations within NetBrain, simplifying governance, rotation, and audit trails for every system NetBrain touches. One place to rotate credentials, one place to audit access — no per-integration credential sprawl. 

Parser Assistant by AI

Convert CLI output into production-ready parser rules in minutes, not hours. Engineers paste raw device output and the AI generates accurate, validated parser rules on the spot. Behind the scenes, the AI produces regex and structured extraction logic — you describe what you want, it handles the syntax. 

• Describe what you want in plain English. AI handles the pattern matching. 
• AI generates validated patterns that work the first time. No debugging. 
• Your network team knows the CLI outputs. That’s all the expertise needed. 

Two Factor Authentication

Enterprise-grade identity security with support for any third-party 2FA provider. IT and security teams enforce multi-factor authentication through any compatible third-party 2FA provider. Meets compliance requirements for privileged access to network infrastructure without custom authentication development.