Install in One Command
A single curl pipes Tink onto any Linux server. No dependencies, no config files, no package manager gymnastics. Thirty seconds to running.
curl -fsSL https://tink.bot/install | shIntroducing
One command to install. Continuous scanning. Plain-English diagnosis. Guided fixes from the CLI. Meet the AI that keeps your infrastructure from quietly falling apart.

“Every server is slowly falling apart. The question is whether anyone is paying attention.”
The Problem
Disks fill up at 3 AM. Memory leaks compound over weeks. Certificates expire on holidays. By the time a human notices, the page is already loading slowly, the queue is backed up, or the service is down entirely.
Monitoring dashboards help — but only if someone is watching them. Alert fatigue is real. Half the notifications get snoozed. The other half require context no dashboard can give you: what actually happened, why it matters, and what to do about it.
This is the gap Tink was built to fill.
The Solution
A single curl pipes Tink onto any Linux server. No dependencies, no config files, no package manager gymnastics. Thirty seconds to running.
curl -fsSL https://tink.bot/install | shTink monitors CPU, memory, disk, services, processes, and network — without you asking. It idles at under 20 MB of RAM. You won't notice it. Your problems will.
No cryptic error codes. No log spelunking. Tink tells you what's wrong in language a human can act on: “Disk /var is 94% full — 12 GB of old journal logs can be safely rotated.”
Tink never runs commands without asking. It proposes a fix, explains why, shows you the exact command, and waits for your explicit approval. Every action is logged for audit.
Talk to your server from your phone. Tink's Telegram integration turns server management into a conversation — ask questions, approve fixes, get alerts, all from a chat window.
One dashboard for every machine. See health status, scan history, open issues, and audit trails across your entire fleet — whether that's one server or fifty.
How It Works
One command. Tink downloads, registers with your account, and starts its first scan automatically.
CPU, memory, disk, services, open ports, certificates, processes — the agent checks everything on a configurable interval and flags anything abnormal.
When something is wrong, Tink doesn't just say “high CPU.” It tells you which process, how long, what the likely cause is, and what you can do about it.
Tink proposes a remediation plan with the exact commands it wants to run. You review, approve, and watch. Nothing executes without your say-so.
Scans, diagnoses, proposed fixes, approvals, and outcomes — a complete audit trail accessible from your dashboard.
“The best sysadmin is the one who never gets paged.”
Trust & Safety
Tink operates on a trust-level system. At the lowest level, it only observes and reports. At higher levels, you can authorize it to take specific autonomous actions — but the default is always to ask first.
Credentials never leave your server
Complete audit trail on every action
Agent ↔ API communication over TLS
Pricing
Free
1 machine · hourly scans
$9/mo
per machine · 5-min scans
$29/mo
per machine · real-time

One command. Thirty seconds. Your server finally has someone watching out for it.
curl -fsSL https://tink.bot/install | sh