Feature

Your Server Speaks AI: Meet Panelica OpsAI — 15 Expert Agents That Actually Execute

March 19, 2026

Back to Blog
Introducing Ops AI

Your Server Speaks AI Now

The first control panel where AI doesn't just suggest — it executes. 15 expert agents, 250 API endpoints, one conversation.

15 Expert Agents 250 API Endpoints 5-Layer Security Full Docker Support

The Problem Every Admin Knows

It's 3 AM. Your monitoring alerts fire. A customer's WordPress site is down. You SSH in, check nginx logs, find a PHP-FPM crash, trace it to a memory limit, adjust the pool config, restart the service, verify the site — and 45 minutes later, you're back in bed. For one site.

What if you could just say: "Check why example.com is down and fix it" — and it actually happens?

Not a Chatbot. An Operator.

It Executes, Not Suggests

Other AI tools tell you what to do. OpsAI does it — creates domains, writes firewall rules, configures services, deploys containers. Real actions, real results.

15 Domain Experts

Switch between specialized profiles — Security Expert knows ModSecurity rules, Email Expert understands DKIM/SPF, WordPress Expert wields WP-CLI. Each one loaded with deep system knowledge.

Isolated & Audited

Runs as a sandboxed user, never as root directly. Every action logged. Every session recorded. You see everything in real-time and can abort instantly.

How It Works

From your browser to your server — in real-time

Browser
xterm.js Terminal
WebSocket
Real-time PTY
Go Backend
250 API Endpoints
Claude AI
Expert Profiles
Your Server
Real Operations

15 Expert Agents, One Panel

Each agent is loaded with deep knowledge of its domain — system prompts, API references, safety rules, and real operational experience.

General

Day-to-day server ops. Health checks, service restarts, disk cleanup, log analysis.

"Check server health and report any issues"
Domain

DNS zones, SSL certificates, virtual hosts, reverse proxies, domain provisioning.

"Add example.com with SSL and PHP 8.4"
Security

IP blocking, Fail2ban, nftables firewall, ModSecurity WAF, ClamAV, AppArmor.

"Run a full security audit on this server"
Email

Postfix, Dovecot, DKIM/SPF/DMARC, spam filters, mail queue, autoresponders.

"Why are my emails going to spam?"
Database

MySQL & PostgreSQL optimization, slow queries, backups, user management.

"Find and optimize slow queries"
WordPress

WP-CLI, plugin management, cache setup, security scans, staging environments.

"Install WordPress with Redis caching"
Docker

Containers, images, compose, app templates, reverse proxy, resource limits.

"Deploy a Redis cluster with Docker Compose"
Backup

Full server backups, snapshots, scheduled jobs, restore, integrity verification.

"Create a full backup and verify integrity"
System

Services, CPU/RAM/disk, cgroups v2, process management, system updates.

"Show resource usage per user with cgroups"
Cloudflare

DNS sync, cache purge, SSL modes, under-attack mode, zone analytics.

"Purge Cloudflare cache for example.com"
Files

Browse directories, manage permissions, compress/extract, disk usage analysis.

"Find the largest files consuming disk space"
SSH

SSH users, key management, session monitoring, SSHD config, jail access.

"Show active SSH sessions and login history"
FTP

FTP users, quotas, active sessions, transfer logs, server settings.

"Create FTP user with 5GB quota"
Cron

Scheduled tasks, execution logs, templates, bulk operations, system cron.

"Schedule daily database backup at 2 AM"
Git

Repositories, branches, deploy keys, webhooks, deployment environments.

"Deploy latest commit to production"

Real-World Scenarios

These aren't demos. These are actual operations OpsAI performs on production servers, every day.

Scenario 1: "We're Under Attack"

3:00 AM — DDoS detected, 50K requests/minute from 200+ IPs
OpsAI — Security Expert
admin ▸ We're getting DDoS'd. Analyze the attack and block it.

OpsAI ▸ Analyzing access logs and connection patterns...
  Identified 247 attacking IPs from 3 ASNs (AS14061, AS16509, AS13335)
  Attack pattern: HTTP flood targeting /wp-login.php and /xmlrpc.php
  Peak rate: 52,000 req/min (normal baseline: 200 req/min)

OpsAI ▸ Taking countermeasures...
  Blocked 247 IPs via nftables firewall
  Added rate limiting rule: 30 req/min per IP on /wp-login.php
  Enabled ModSecurity OWASP CRS paranoia level 2
  Activated Cloudflare Under Attack mode for affected domains
  Created Fail2ban jail: wordpress-flood (bantime: 24h)

Attack mitigated. Traffic dropped from 52K to 180 req/min. All sites operational.
45 min
Manual response time
90 sec
OpsAI response time
5 actions
Coordinated simultaneously

Scenario 2: "My WordPress Is Slow"

Customer ticket — page load time 8.2 seconds
OpsAI — WordPress Expert
admin ▸ customer.com is loading in 8 seconds. Diagnose and fix it.

OpsAI ▸ Running WordPress performance audit...
  PHP OPcache: disabled (every request recompiles PHP)
  Object cache: none (87 DB queries per page load)
  3 plugins causing slow queries (avg 1.2s each)
  Unoptimized images: 23 images totaling 18MB on homepage

OpsAI ▸ Applying optimizations...
  Enabled OPcache (128MB, 10K files, revalidate 60s)
  Installed Redis object cache (wp redis enable)
  Configured PHP-FPM pool: pm=ondemand, max_children=15
  Deactivated problematic plugins, suggested alternatives
  Enabled Cloudflare Polish + Mirage for image optimization

Page load time: 8.2s → 1.1s | DB queries: 87 → 12 | TTFB: 3.4s → 0.18s

Scenario 3: "Set Up a New Customer"

Complete hosting environment from zero — one sentence
OpsAI — General Expert
admin ▸ New customer: newclient.com, WordPress site, business plan, email needed.

OpsAI ▸ Provisioning complete hosting environment...
  Created user "newclient" with Business plan (10GB disk, 100GB BW)
  Added domain newclient.com with PHP 8.4
  DNS zone configured (A, CNAME www, MX, SPF, DMARC)
  SSL certificate issued via Let's Encrypt
  WordPress installed with starter theme
  Redis object cache configured
  Email: [email protected] created (1GB quota)
  FTP access: [email protected] ready
  Cgroup limits applied (2 CPU cores, 2GB RAM)
  PHP-FPM per-user pool active

Full environment ready in 52 seconds. Credentials sent to admin panel.

One sentence created: a Linux user with cgroup isolation, a domain with DNS and SSL, WordPress with caching, an email account with DKIM/SPF, FTP access, and per-user PHP-FPM pool. 10 operations that normally take 30+ minutes — done in under a minute.

Scenario 4: "Emails Going to Spam"

Gmail rejecting with 5.7.26 — SPF/DKIM alignment failure
OpsAI — Email Expert
admin ▸ Emails from client.com are going to spam on Gmail. Fix it.

OpsAI ▸ Running email deliverability audit...
  SPF record: missing -all (using ~all — soft fail)
  DKIM: key exists but using sha1 (Gmail requires sha256)
  DMARC: p=none (Gmail 2025+ requires p=quarantine minimum)
  PTR record: points to generic hostname, not mail.client.com
  TLS: properly configured

OpsAI ▸ Fixing all issues...
  Updated SPF to v=spf1 ip4:x.x.x.x -all (hard fail)
  Regenerated DKIM key with sha256, rotated selector
  Set DMARC to p=quarantine with rua/ruf reporting
  Synced all DNS records to Cloudflare
  Sent test email — delivered to Gmail inbox

Email deliverability restored. All 4 issues fixed in 40 seconds.

Scenario 5: "Disk 95% Full — Server About to Crash"

Critical alert — services failing, MySQL can't write
OpsAI — System Expert
admin ▸ Disk is 95% full. Find what's eating space and clean it up safely.

OpsAI ▸ Analyzing disk usage...
  /var/logs/ — 12.4 GB (old rotated logs not cleaned)
  /var/backups/ — 45.2 GB (3 months of daily backups)
  MySQL binlogs — 8.7 GB (expire_logs_days not set)
  /tmp/ — 3.1 GB (stale session files)

OpsAI ▸ Cleaning safely (keeping latest backups)...
  Cleaned rotated logs older than 7 days → freed 11.2 GB
  Removed backups older than 30 days (kept last 4) → freed 38.1 GB
  Set MySQL expire_logs_days=7, purged old binlogs → freed 7.8 GB
  Cleaned /tmp stale files → freed 3.1 GB

Freed 60.2 GB total. Disk now at 42% usage. MySQL restarted successfully.

Scenario 6: "Deploy My App with Docker"

Node.js app + PostgreSQL + Redis + Nginx reverse proxy
OpsAI — Docker Expert
admin ▸ Deploy myapp.com: Node.js app at /home/user/myapp, needs PostgreSQL and Redis, with SSL.

OpsAI ▸ Building Docker Compose stack...
  Generated docker-compose.yml (app + postgres:16 + redis:7)
  Created .env with secure random passwords
  Built app image from Dockerfile (Node.js 22 Alpine)
  Containers started: myapp-web, myapp-db, myapp-redis
  Health checks passing on all 3 containers
  Nginx reverse proxy configured (port 3000 → myapp.com)
  SSL certificate issued for myapp.com
  Cgroup limits applied (4 CPU, 4GB RAM per container group)

Full stack deployed. https://myapp.com is live with SSL.

Scenario 7: "Migrate 50 Sites from cPanel"

Full migration with zero downtime — domains, databases, emails, SSL
OpsAI — General Expert
admin ▸ Migrate all sites from old cPanel server (203.0.113.50). Here are the SSH credentials.

OpsAI ▸ Connecting to source server and discovering sites...
  Found 50 domains, 47 databases, 63 email accounts, 28 SSL certs
  Total data: 124 GB files + 18 GB databases

OpsAI ▸ Migrating... [████████████████████░░░░] 78% — site 39/50
  Users created with original passwords preserved
  Files transferred via rsync (delta sync, 2.1 GB/min)
  MySQL databases imported with hash-preserved credentials
  Email accounts migrated with IMAP folder structure
  DNS zones replicated, SSL certificates re-issued
  wp-config.php and .env files updated automatically

50 sites migrated. 0 failures. Total time: 47 minutes. Verification passed on all domains.

Scenario 8: "Run a Complete Security Audit"

Comprehensive server hardening check — firewall, malware, config, access
OpsAI — Security Expert
admin ▸ Run a complete security audit on this server.

OpsAI ▸ Running comprehensive security audit...
  Firewall: nftables active, 847 rules, no open unnecessary ports
  Fail2ban: 12 jails active, 34 IPs currently banned
  ModSecurity: OWASP CRS 4.x, paranoia level 1, 0 false positives today
  ClamAV: Scan complete — 0 threats found in 142,847 files
  SSL: 2 certificates expiring within 7 days
  SSH: Key-only auth, non-standard port, no root password login
  PHP: disable_functions set, open_basedir per-user enforced
  Permissions: All home dirs 700, no world-writable files found

OpsAI ▸ Auto-fixing found issues...
  Renewed 2 expiring SSL certificates
  Security score: 96/100 (up from 91)

Audit complete. 8 categories checked. 1 issue found and auto-fixed.

Before & After OpsAI

WITHOUT OpsAI
SSH into server
Check nginx error logs manually
Check PHP-FPM status
Edit pool config with vim
Restart PHP-FPM service
Check MySQL slow query log
Run EXPLAIN on queries
Edit my.cnf for optimization
Install Redis manually
Configure wp-config.php
Test, debug, repeat...
35+ minutes
12 steps, context switching, typo risk
WITH OpsAI
"WordPress on customer.com is slow. Fix it."

That's it. OpsAI:
Diagnoses the root cause
Checks all related services
Applies optimizations
Installs caching layers
Verifies the fix
Reports the results

60 seconds
1 sentence, zero context switching

5-Layer Security Model

Power without compromise — every action controlled, every session recorded

ROOT Only
Only root admins can access OpsAI
Sandbox
Isolated user, isolated directory, isolated credentials
Permissions
Whitelist/blacklist for commands and paths
Audit Trail
Every action logged, every session recorded
Live Oversight
Real-time terminal visibility, instant abort

A Day with OpsAI

09:00 — Morning Health Check System Expert

"Check server health" → CPU 23%, RAM 4.2/31GB, disk 42%, all 20 services running, 0 failed logins overnight.

09:15 — New Client Setup General Expert

"Set up newclient.com with WordPress and email" → Full environment provisioned in 52 seconds.

11:00 — Performance Complaint WordPress Expert

"customer.com is slow" → Diagnosed: no caching, unoptimized queries. Fixed with Redis + OPcache. 8.2s → 1.1s.

14:00 — Docker App Deployment Docker Expert

"Deploy staging environment for the Node.js app" → Compose stack with PostgreSQL, Redis, Nginx proxy — live in 90 seconds.

16:30 — End-of-Day Security Audit Security Expert

"Run security audit" → Firewall, Fail2ban, ClamAV, SSL, SSH — all green. 2 expiring certs auto-renewed. Score: 96/100.

17:00 — Backup & Done Backup Expert

"Create full server backup" → Snapshot created, integrity verified, stored securely. Server managed for the day — in conversations, not commands.

OpsAI vs. The Rest

Feature Traditional Panel ChatGPT + SSH Panelica OpsAI
Executes server commands Manual clicks Copy-paste Direct execution
Knows your server context No Full context
Multi-step automation 1 step at a time Suggests steps End-to-end
Security sandboxing N/A None 5-layer model
Specialized experts Generic 15 domain experts
Audit trail Basic logs None Full recording
Works with your panel Is the panel Separate tool Built-in

AI That Amplifies, Not Replaces

OpsAI doesn't replace system administrators — it gives them superpowers. You provide the strategy and judgment. OpsAI handles the execution at machine speed. The human stays in control, always.

250
API Endpoints
15
Expert Agents
20+
Services Managed
5
Security Layers

Ready to Talk to Your Server?

OpsAI is available now in Panelica. Open your panel, navigate to Tools → Ops AI, and start managing your server with natural language.

Panelica OpsAI is powered by Claude AI from Anthropic. Available on all Panelica installations with an active license.

Share:
Security, built-in.