Every Feature, Explained
From Ghost Text autocomplete to multi-device debugging, see exactly what RTRGRD brings to your workflow.
Ghost Text Autocomplete
Start typing and watch the terminal complete your command. Powered by Learn As You Go (LAYG). No training required.
- Real-Time Learning: Commands you use appear in suggestions
- Device-Scoped: Filtered by the device you're connected to
- OS-Aware: Only shows valid commands for Cisco, Aruba, Juniper, etc.
- <100ms Response: Local FTS + vector search
AI Copilot
Ask questions in natural language. Get verified commands back. Drag them directly into your terminal.
- Sidebar Chat: Full conversation with context awareness
- @ai Inline: Query directly from the terminal prompt
- 35,000+ Verified Commands: Grounded in real CLI documentation
| Query Type | Model Tier | Response |
|---|---|---|
| Simple lookup | Fast model | <2 sec |
| Troubleshooting | Standard model | 3-5 sec |
| Deep reasoning | Expert model | 5-10 sec |
| BYOK (Your API Key) | Your provider | You pay providers directly |
Mission Control
Force-graph topology visualization with LLDP auto-discovery. See your network. Ask questions about it.
- LLDP/CDP Discovery: Automatic neighbor detection
- Dynamic Animations: Traffic flow visualization
- AI Topology Awareness: "What's the path from Spine1 to Leaf12?"
- Workflow Targets: Select nodes for bulk operations
War Room
Debug up to 4 devices simultaneously with AI-guided playbooks. Correlate logs. Find root cause.
| Mode | Use Case |
|---|---|
Guided |
AI-led interactive debugging |
Health Check |
Standard diagnostics across devices |
Playbook |
Execute predefined troubleshooting scripts |
- Severity Rating: Critical, Warning, Info, OK findings
- Suggested Fixes: AI-generated remediation commands
Multi-Device Command Deck
Plan, scope, and execute changes across multiple devices in parallel, with hallucination-resistant guardrails baked into every step.
- SCOPE TOKEN: A signed plan boundary; commands outside the agreed scope are rejected before they reach the wire
- AMENDMENT MODE: Edit a previously-executed plan with a clean diff and re-validation before re-running
- AI Best-Guess Prefix: Every AI-generated command is labeled with confidence and grounding source
- Per-OS Syntax Grounding: Per-device scaffolds validate against the device's actual CLI grammar, not a generic vendor doc
- Checkpoint Safety: Automatic rollback preparation across the fleet
| OS | Checkpoint Method |
|---|---|
| Cisco NX-OS | commit confirmed |
| Cisco IOS-XE | revert timer |
| Arista EOS | commit timer |
| Juniper Junos | commit confirmed |
| Aruba CX | checkpoint auto confirm |
| 20+ others | Snapshot rollback |
Sanitization Layer
30+ credential patterns blocked before any cloud AI request. Watch the Guardian Orb for real-time protection status.
| Pattern | Example Match |
|---|---|
| Passwords | password 7 **** |
| SNMP Strings | snmp-community **** |
| API Keys | bearer **** |
| Crypto Material | crypto key **** |
| TACACS/RADIUS | tacacs-server key **** |
- Locally Encrypted: Credentials stored with OS-level encryption
- Privacy Mode: 100% local inference with Google Gemma 4 (E2B/E4B/26B MoE), sized to your machine
20+ OS Types Supported
Ready to Try?
15-day full-AI trial, then keep the free tier forever. No credit card required. Windows 10/11.
Join the Waitlist