rConfig Iris
rConfig V8 AI Center Pro Standard and above rconfig.com/ai

Ten AI capabilities.
One governed place.

Iris is the AI built into rConfig. It reads your network configs and explains them in plain words. It lets the AI tools your team already uses read the same facts. It never changes a device.

10AI capabilities
7read-only MCP tools
13AI providers
0per-prompt fees
Inside rConfig
01

Config Analysis

Ask about any config in plain English.

“Which edge routers allow SSH from anywhere?”
02

Drift Narration

See what changed, and whether it matters.

“This BGP change could leak transit routes.”
03

Policy and Compliance

Know why a check failed. Draft the next rule.

“The Telnet rule failed. A CIS rule is drafted for review.”
04

Optimization

Tune rConfig itself, straight from the official docs.

“For 5,000 devices, run 8 to 12 workers.”
Iris · ask your network Read-only
We have a PCI audit coming. Across all edge routers, which ones allow management access from outside our 10.0.0.0/8 range?
I checked 48 edge routers. 6 allow management access from outside that range. The worst is edge-rtr-11: SSH is open to any address, and Telnet is still on. Here are the lines that prove it.
edge-rtr-11 · running-config
line vty 0 4
  transport input telnet ssh
  access-class 23 in  ! permits any
6 of 48 exposed Sorted by risk Proof for every device
Example answer. Iris explains and suggests. Your engineers decide. Your model · your keys
For your AI tools
05

MCP Server

Your AI agents read live network facts. They can look. They cannot touch.

› find switches hit by the new CVE
· rconfig · search_configs
  34 devices matched
· jira · create_issue × 12
  tickets raised, proof attached
06

Agent Skills

Teach any agent how rConfig works, so the job is right first time.

Claude Code Codex Cursor Gemini CLI
07

API-ready docs

The REST API as Markdown your agent can read. Real calls, no guessing.

GET /api/v2/devices  → docs.md
On your terms
08

Bring your own model

Use the AI you already trust. In the cloud, or fully inside your own network.

Anthropic OpenAI Azure OpenAI AWS Bedrock Google Gemini Mistral Cohere DeepSeek Groq xAI OpenRouter Ollama · local OpenAI-compatible · local
09

Governance and audit

Your roles decide who uses AI. Every question is on record.

a.okafor · Config Analysis · 09:12
s.nolan · Drift Narration · 09:09
10

Privacy by design

Each capability sends only what it needs. The list is published.

With a local model, nothing leaves your network.
Network AI maturity model Four steps. We tell you which one we are on.
1 · Assisted InsightrConfig is here today
2 · Augmented OperationsTuned to your standards
3 · Supervised AutomationAI proposes, a person approves
4 · Autonomous OperationsThe horizon
rconfig.com/ai/maturity-model