A modern TUI for
Managing Servers
Manage AWS, OVH Cloud, and Hetzner Cloud servers from one TUI — SSH, SCP, full server lifecycle, AI log analysis, and an MCP server for agents like Claude Code.
Servonaut AI — hosted chat & server triage included on Solo & Teams plans.
$ servonaut # Launch TUI
Servonaut AI Assistant
MCP Server OnlineCommand your Infrastructure
One TUI for AWS, OVH Cloud, Hetzner Cloud, and any custom server — with an AI-callable MCP surface on top.
Multi-cloud lifecycle
AWS EC2, OVH (Public Cloud / VPS / dedicated), Hetzner Cloud, and any custom SSH host in one searchable list. Create, start, stop, reboot, and delete from the same Manager screen.
> + custom SSH
MCP server & hosted AI
46 MCP tools your agent can call — full provider lifecycle, server memory, CloudWatch & CloudTrail analysis, IP-ban controls, run commands, fetch logs. Hosted Servonaut AI on Solo & Teams plans, or bring your own Anthropic / OpenAI / Gemini / Ollama key.
Security & observability
CloudTrail audit, CloudWatch log analysis, IP ban via AWS WAF / Security Groups / NACLs, and AbuseIPDB enrichment — all wired into the same instance row.
> IP Ban Manager
AI-Powered Server Management
Give your AI agents direct access to your infrastructure via the Model Context Protocol.
Servonaut ships with a built-in MCP server. One command installs it into Claude Code, giving AI agents read and write access to your servers through structured, auditable tool calls.
Works with Claude Code, Cursor, Windsurf, and any other MCP-compatible client. Guard levels (readonly, standard, dangerous) let you control exactly what the AI can touch.
Surface errors and anomalies across all your servers in seconds.
Browse CloudTrail events and manage WAF rules from a guided flow, with every change audited.
Describe what you want. The AI figures out which tools to call.
→ Calling get_logs(server="prod-1", lines=100)
→ Found 3 critical errors in nginx.log
→ Analyzing patterns...
> Deploy the fix to staging
→ Calling run_command(server="staging", cmd="git pull && restart")
→ Deployment complete. Service healthy.
> Scan CloudTrail for suspicious activity
→ Calling cloudtrail_lookup_events(hours_back=24)
→ 2 unusual IAM events detected from 203.0.113.42
→ Calling ip_ban_set(ip_address="203.0.113.42", config_name="waf-blocklist", action="ban")
→ IP banned. WAF rule created.
Up and running in minutes
No agents to configure, no infrastructure to provision. Just install and go.
Install
One command. No dependencies to wrangle. Works on Linux, macOS, and Windows.
Configure
Point to your accounts. AWS EC2 auto-discovers from the AWS CLI profile. OVH and Hetzner use one-shot setup wizards. Custom SSH hosts slot in via ~/.servonaut/config.json.
Manage
SSH, deploy, create & destroy servers, analyze logs, audit CloudTrail, manage WAF rules — all from your terminal. Or let your AI agent handle it via 46 MCP tools.
Built for modern infrastructure teams
Ready to level up?
The CLI is free and open source. Upgrade for cloud config sync and a hosted MCP server for your AI agents.
Quick Installation
curl -sSL https://raw.githubusercontent.com/zb-ss/ec2-ssh/master/install.sh | bash
# 2. Windows (PowerShell)
irm https://raw.githubusercontent.com/zb-ss/ec2-ssh/master/install.ps1 | iex
# 3. Pipx (Recommended for isolated install)
pipx install servonaut
# Optional AI and MCP dependencies
pipx inject servonaut httpx mcp
Requires Python 3.10+, an SSH client, and AWS CLI configured.