Conversational Infrastructure, Digital Twins & Long-Term Memory

The RasnuOS conversational interface lets you provision, troubleshoot, and optimize single servers or multi-cloud fleets in plain English—without writing brittle bash scripts or clicking through dozens of nested settings menus.

Unlike a generic chatbot that knows nothing about your servers until you paste terminal output into it, RasnuOS equips every conversation with four core capabilities:

1. Live @server and @/path/to/file Mentions + File Attachments
2. Real-Time Server Digital Twins (automatic live OS, metric, port, and container awareness)
3. Long-Term Institutional Memory (remembering your custom paths, ports, and architecture across sessions)
4. Interactive Generative UI Cards (inline metric cards, sortable data tables, charts, file trees, and action lists)

---

1. Targeting Servers, Remote Files & Attachments

@server_name Fleet Autocomplete

Typing @ anywhere in the chat composer opens an instant autocomplete menu of all servers and server groups in your organization:

• Single-Server Precision: Mentioning a single server (for example, "Check why disk usage is climbing on @prod-db-01 and clean up old rotated logs") locks the conversation context onto that specific host.
• Multi-Server Orchestration: Mentioning two or more servers in a prompt (for example, "Stream the analytics PostgreSQL database from @db-eu-01 to @db-us-01 over P2P and verify row counts match") automatically loads the live Digital Twins for both servers and activates multi-server orchestration.

@/path/to/file Remote Filesystem Mentions

Typing @/ in the composer queries the active server's filesystem in real time, allowing you to autocomplete and reference remote configuration files or logs directly inside your prompt:

"Inspect @/etc/nginx/sites-available/api.conf and @/var/log/nginx/error.log on @edge-01 and fix the upstream 502 Bad Gateway timeout."

Drag-and-Drop Multimodal Attachments

You can attach architecture diagrams, terminal screenshots, configuration files, or log snippets directly to any message. RasnuOS analyzes both visual attachments (such as a screenshot of a Grafana spike or a cloud VPC diagram) and text attachments alongside your server's live telemetry.

---

2. The Server Digital Twin (Zero-Prompt Context)

When you open a conversation with RasnuOS, you never have to waste time telling the assistant "This server runs Ubuntu 24.04 with 8 CPU cores, 16 GB of RAM, Nginx on port 443, and PostgreSQL on port 5432."

Before the AI generates its very first response, RasnuOS automatically injects a live Digital Twin of your target server compiled from the Agent's real-time 15-second telemetry and service discovery:

Rendering diagram…

What the Digital Twin Includes Automatically

• Operating System & Hardware Profile: OS family (Linux, Windows Server, macOS), distribution release (Ubuntu 24.04 LTS, Debian 12, Windows Server 2022), CPU architecture (amd64 / arm64), and logical CPU core count.
• Live Resource Headroom: Current CPU %, RAM %, Disk %, Swap %, 5-minute Load Average, Disk IO Wait %, and any recent kernel OOM Kill events.
• Discovered Services & Listening Ports: Every active daemon (nginx, postgresql, mysql, redis-server, mongod, etc.), its current status (running, failed, stopped), and its exact network bind address (127.0.0.1:5432 vs. 0.0.0.0:443).
• Docker Container Inventory: All running and exited containers, image tags, health statuses, and host port mappings.
• Recent Incident Context: Any active or recently resolved AutoPilot incidents on that host.

Smart Token Optimization ("Single-Server Prompt Diet")

If your organization manages 100 servers, loading the full Digital Twin of all 100 servers into every chat message would waste tokens and slow down response times. When you are working on a single server, RasnuOS loads the deep Digital Twin for that active server while condensing the rest of your fleet into a lightweight 1-line-per-server directory (Name | IP | OS | Status)—keeping responses fast while preserving full fleet awareness.

---

3. Long-Term Infrastructure Memory & Rolling Summaries

Infrastructure knowledge should compound over time. If you deploy a custom microservice to /opt/services/ledger-engine running on custom port 9410 on Monday, you should not have to remind RasnuOS where that service lives three months later.

RasnuOS maintains two layers of persistent memory for every server in your organization:

1. Semantic Long-Term Server Memory

Whenever a conversation completes, RasnuOS automatically extracts durable architectural facts and troubleshooting lessons learned about that server (such as custom application directories, non-standard ports, custom systemd unit names, and runtime versions) and indexes them into your organization's isolated Semantic Vector Memory:

Rendering diagram…

• Automatic Deduplication (>= 92% match): Prevents storing repetitive copies of facts the system already knows.
• Automatic Stale-Fact Supersession (82%–92% match): If a service moves from port 8080 to port 8081, RasnuOS automatically supersedes and replaces the outdated fact so future conversations always use the latest port.

2. Five-Section Rolling Session Compression

During long, multi-hour debugging or deployment sessions with dozens of command steps, RasnuOS automatically compresses older turns into a structured 5-Section Operational Summary so the assistant never loses track of earlier progress:

GOAL:  Migrate the billing API to Node.js 22 with zero downtime behind Nginx.
DONE:  Installed Node.js 22, cloned release v4.2.0, verified build artifacts.
STATE: Systemd unit billing-api.service is active on 127.0.0.1:4000; Nginx is running.
PATHS: App root: /opt/billing-api | Nginx config: /etc/nginx/sites-available/billing.conf
NEXT:  Switch Nginx upstream port from 3000 to 4000 and run health verification.

---

4. Interactive Generative UI Components

Nobody wants to squint at 200 lines of raw ss -tulnp or df -h text output when making production decisions. Alongside terminal logs and markdown explanations, RasnuOS renders Interactive Visual Components directly inline inside the conversation stream:

Visual Component	When RasnuOS Renders It	What You Can Do With It
Metric Cards (MetricCard)	Server health checks, database connection pools, SSL countdowns, resource summaries	View color-coded status indicators (healthy, warning, critical) and key-value KPIs at a glance.
Data Tables (DataTable)	Open port audits, top CPU/RAM processes, compliance scan results, Docker container lists	Inspect structured rows and columns cleanly without parsing raw terminal spacing.
Visual Charts (UiChart)	Disk usage breakdowns by directory, memory allocation comparisons, performance trends	Compare resource distribution visually via inline bar, line, or doughnut charts.
Directory Trees (FileTree)	Exploring application folders (/var/www/app), Nginx config hierarchies, or backup directories	Browse nested folders, file sizes, and directory structures interactively.
Action Checklists (ActionList)	Recommended security hardening steps, post-deployment follow-ups, optimization options	Trigger recommended follow-up actions with a single click.

Example Generative UI Payload Rendered in Chat

{
  "render_ui": [
    {
      "type": "MetricCard",
      "props": {
        "title": "PostgreSQL 16 Health (@prod-db-01)",
        "metrics": [
          { "label": "Service Status", "value": "Active (Running)", "status": "healthy" },
          { "label": "Active Connections", "value": "42 / 200", "status": "healthy" },
          { "label": "Shared Buffers", "value": "4.0 GB", "status": "info" },
          { "label": "Replication Lag", "value": "0 ms", "status": "healthy" }
        ]
      }
    },
    {
      "type": "UiChart",
      "props": {
        "title": "Disk Usage Breakdown (/var)",
        "chartType": "bar",
        "labels": ["/var/lib/postgresql", "/var/lib/docker", "/var/log", "/var/cache"],
        "datasets": [
          {
            "label": "Size (GB)",
            "data": [28.4, 14.2, 3.1, 1.8]
          }
        ]
      }
    }
  ]
}

---

5. What a Server Digital Twin Snapshot Looks Like

Below is an example of the live Digital Twin context RasnuOS assembles automatically for your active server on every turn:

=== ACTIVE SERVER DIGITAL TWIN: prod-api-01 ===
OS: Ubuntu 24.04.1 LTS (linux / amd64) | CPU Cores: 8 | Private IP: 10.0.1.20
Live Telemetry (Updated 5s ago):
- CPU: 18.4% | RAM: 62.1% (9.9 GB / 16.0 GB) | Disk: 41.2% | Swap: 0.0%
- 5m Load Average: 1.15 | Disk IO Wait: 0.8%
Discovered Services:
- nginx (running)        -> 0.0.0.0:80/tcp, 0.0.0.0:443/tcp
- api-worker (running)   -> 127.0.0.1:8080/tcp
- redis-server (running) -> 127.0.0.1:6379/tcp
- postgresql (running)   -> 127.0.0.1:5432/tcp
Docker Containers:
- metrics-exporter (prom/node-exporter:latest) -> Up 21 days (127.0.0.1:9100->9100/tcp)

---

6. High-Leverage Conversational Prompt Examples

Engineering Task	Example Prompt to Type in Conversations
Zero-Downtime App Rollout	"Deploy the main branch of https://github.com/acme/billing-api to /opt/billing on @prod-api-01, configure a systemd service on port 8080, and expose it via Nginx on billing.acme.com with Let's Encrypt SSL."
Root-Cause Outage Triage	"Users are seeing intermittent 504 Gateway Timeouts on @edge-eu-01. Inspect @/var/log/nginx/error.log, check upstream response latencies, and show the top bottleneck processes in a table."
Cross-Cloud Database Migration	"Stream the customers_prod PostgreSQL database from @legacy-aws-db directly to @hetzner-db-01 over an encrypted P2P tunnel and verify that all tables and indexes match."
Fleet-Wide Security Hardening	"Audit SSH configuration and open ports across @web-prod-01, @web-prod-02, and @db-prod-01. Disable SSH password login and block public access to port 5432."