$ opsterm "what can you do?"

OpsTerm

A local AI terminal assistant that lives in your terminal.
SSH into any server without leaving your AI behind.
⚡ Zero dependencies · Any AI provider
🗜️ RTK AI — 60-95% token compression 🔗 Multi-hop SSH
⚡ What it looks like
~/opsterm — bash
# Ask AI for a command
$ opsterm how to find large log files
$ find /var/log -type f -size +100M -exec ls -lh {} \;

# SSH without remembering IPs
$ opsterm ssh vps-utama
🔑 Connecting to vps-utama...
ubuntu@43.157.204.199:~$


# Explain any command output
$ docker ps | opsterm "any errors?"
✓ All 3 containers healthy. No errors found.

⚡ Quick Install — one command, no repo

# 1. Download & run (single file, zero deps) $ curl -L https://raw.githubusercontent.com/edsuwarna/opsterm/main/bin/opsterm -o ~/.local/bin/opsterm $ chmod +x ~/.local/bin/opsterm $ opsterm init # 2. Add your AI provider $ opsterm provider add default --api-key 'sk-...' --model gpt-4o $ opsterm --help
✨ Features
🤖

AI Chat

Ask anything — commands, explanations, scripts. Works with any OpenAI-compatible provider.

🔑

Smart SSH

SSH without IPs. Fuzzy matching, multi-hop via --via, YAML config.

Workflows

Multi-step YAML automation: SSH → SCP → Local commands, all in one file.

🔐

Vault

AES-128 + PBKDF2 encrypted credential store. Zero external deps.

🔗

Pipe Mode

cmd | opsterm "analyze this" — explain output, detect errors, get insights.

⌨️

Tab Completion

Bash + Zsh auto-complete for servers, workflows, and subcommands.

📁

SCP Transfer

Upload/download files via server name: opsterm scp file.txt server:/path

📋

History

SQLite-backed history. Revisit, explain, or re-run past commands.

🗜️

RTK AI

Auto-compress command output 60-95% with RTK before AI. Zero config — if installed, it just works.

🏗️ Under the hood
Single Python file — zero deps
Config at ~/.opsterm/
Any OpenAI-compatible provider
Multi-hop SSH via ProxyJump
Auto-migrate from ~/.ai-workflows
Shell operators in local commands
RTK AI — 60-95% token compression
📚 Documentation