Get tasks doneBusiness size: Solo

A read-only Claude Code agent checked ten machines and found a server with no backup job on day one

Sapuan (Romzi) · Malaysia · Claude Code, Proxmox, SQLite

Go to the skill

What the agent did

He runs two Proxmox hosts, a NAS, a media server and five internet servers for his websites. Claude Code runs on its own small virtual machine and pulls jobs from his home dashboard, which offers 15 ready-made, read-only checks such as disk space, failed services, waiting updates and "What should I do first this week?". Each check runs now or on a schedule, and every job must finish before his 07:30 morning report.

Result, as claimed

In the first few hours it reported a monitoring service that had quietly failed on one internet server, listed security updates waiting on three machines in a sensible order, and pointed out that his main Proxmox host had no backup job at all, so he set one up. It also raised one false alarm, which he traced to his own fragile check.

This is the result as claimed in the public source linked below. It has not been checked and is not a typical outcome.

Proof

Read the write-up on his blog

Try it yourself

Give an AI a fixed menu of read-only checks for your servers or website hosting, so it reports problems without changing anything.

Difficulty: Advanced

Prompt

I run [number] servers for [my websites or business]. Help me set up a read-only health check menu for an AI agent. Suggest 10 to 15 checks in plain words, grouped into daily health, security, housekeeping and plain-language summaries. For each check, give the read-only command it would use, how often to run it, and what counts as OK, WARN or CRIT. Every answer should start with OK, WARN or CRIT, then a short table, then what needs my attention with the exact command I could run myself. The agent must never run a change.

Tools

  • Claude Code Named in the write-up. Runs on its own VM and writes each check's answer
  • Proxmox VE His virtual machine hosts, one of the systems it checks

Steps

  1. List the machines you run and what each one does.
  2. Run the prompt to get your check menu.
  3. Give the agent a separate read-only login on each server, key only.
  4. Start with one daily report and read the first answers closely.

Watch out: An AI can misread a stale log or a fragile check, so treat its first reports as a draft and keep it unable to change anything.

Starter version by Solopreneur Code

Skill: Build with a coding agent

Add this skill

Build with a coding agentDifficulty: Advanced

Read-only server health menu

Designs a fixed menu of read-only checks for the servers behind your websites, with OK, WARN or CRIT answers and no changes made.

Works with
Claude, ChatGPT, Gemini Any AI can design the menu. Running the checks on a schedule needs a coding agent such as Claude Code with read-only access to your servers.
Tools in the story
Claude Code, Proxmox VE
Country
Malaysia

How to use this: Copy the skill, paste it into Claude, ChatGPT or Gemini, and ask it to save it as a skill. If it cannot, it will still follow the steps.

Show skill

If the Copy button does not work, select this text and copy it.

Job: check and fix computers and code

Other tools for this job

Tools used

  • Claude CodeFree, then Pro $20 a month ($17 a month billed yearly), Max $100 and Ultra $200 a month. Claude Code is included in paid plansChecked 2 November 2026. Price source: Claude Code, opens in a new tab
  • ProxmoxSee official pricing
  • SQLiteSee official pricing

Could also do this

Sign up for the newsletter

Get the next story from Solopreneur Code in your inbox.

Form not loading? Subscribe on Substack

Back to all stories

More like this

Browse all stories
Get tasks doneBusiness size: SoloIncludes prompt

A newsletter writer's Claude Code agent turned screenshots into 38 scheduled Substack Notes he did not write, with no engagement lift

Dheeraj Sharma @genaiunplugged · Country unknown · Claude Code, SubflowAI Substack connector, macOS scheduler, Substack Notes

Each morning a headless Claude Code run takes the oldest screenshot, classifies it, writes a 45 to 115 word Note in his voice, checks it against copy gates, uploads the image, schedules it for the next free day, reads the schedule back and archives the screenshot. Screenshots with emails, phone numbers or orders are set aside.

Result, as claimed

29 Notes scheduled by 23 Sep and 38 by 12 Oct with none written by him. 12 measured Notes had a median of 3.5 likes against 5 for his other Notes, which he reads as no lift.

Get tasks doneBusiness size: SoloIncludes prompt

A Fukuoka developer had Claude Code build a bookkeeping pipeline that reconciled 200 transactions and flagged its own doubts

Ken Imoto @kenimo49 · Japan · Claude Code, MCP connection to his accounting platform (platform not named)

Built a re-runnable pipeline that pulls the month's transactions over MCP, matches bank rows to receipts by amount and date, proposes a category with a one-line reason for each row and flags rows it isn't sure about.

Result, as claimed

200 transactions reconciled and 11 miscategorised, a 94.5% hit rate. Review shrank to 18 flagged rows plus a sample. 7 of the 11 errors were in the flagged pile and 4 were caught on the sample.

Get tasks doneBusiness size: CompanyIncludes prompt

A Malaysian point-of-sale company runs about 170 AI agents its co-founder built, and sales admin per lead fell from about 15 minutes to about 1

Fong Wai Hong · Malaysia · Claude Code, Salesforce

When someone books a demo, an agent researches the shop and briefs the caller. Other agents rank every open lead into a live call list that re-sorts about every 30 seconds, so a promised callback rises to the top. After each call the conversation is transcribed and the lead record is filled in. He also replaced a couple of hundred Zapier automations with his own stack in about four weeks.

Result, as claimed

Fong says the average time to close out a lead is now about 1 minute, down from about 15. He says reps make two times more calls a day and meetings booked per 100 leads went up about 25%, which he says shows up in the bottom line. He says the marketing team used to spend 30 to 40% of its time fixing broken Zaps.