Back to jiejuefuyou.github.io
First 10 clients only — Starter $99 (will be $199 after spot 10). 3/10 spots taken.

Claude Code 24/7 - I Teach You How

A 150-line Python script plus 5 markdown templates that lets a Claude Code agent run continuously for 36-plus hours, outputting code commits or article drafts every hour -- for 20 USD/month (Pro subscription). No API billing. No VSCode open. I built it in 14 days. Now I teach you in 30 minutes.

Why Claude Code Cannot Run 24/7 Out of the Box

Root cause: Claude Code is an "IDE assistant" architecture -- not an "agent runtime." Close VSCode and it stops. Its CronCreate tasks die with the window. Memory evaporates when the session ends.

Most people hit one of these walls:

  • VSCode-closed = agent-dead -- "24/7 autonomy" is marketing, not reality
  • API costs explode -- switching to API billing = 3000-plus CNY/month (vs 150 CNY/mo Pro subscription)
  • GitHub Actions kill context -- each run starts fresh, consuming 5-10x more tokens
  • PyAutoGUI steals focus -- fires抢窗口, disrupts your active work, screen flashes
  • Agents override each other -- continuous firing causes loops that overwrite prior work

The fix: Layer an orchestration system outside Claude Code. This is what Anthropic will not give you -- and it is the difference between "I tried agents" and "agents shipped my product."

What This System Actually Shipped

14-Day Real Case Study (April 2026)

4
iOS Apps shipped
11
GitHub repos
9
dev.to articles
20 USD
monthly cost

All verified on GitHub @jiejuefuyou. Privacy Manifest passed. Zero analytics SDKs. Zero network calls from apps.

View GitHub
Memory layer 4 files (MEMORY.md / user_persona.md / autonomy_contract.md / token_economy.md) -- agent re-enters any session and immediately knows context.
State + decisions state.yml (multi-app status) plus decisions.md (9-plus ADR logs) -- agent never re-litigates prior choices.
Desktop loop 150-line Python GUI automation -- idle detection, pause lock, 3-layer busy detection, focus restoration. Runs with VSCode closed.
Token economy Active / Conservative / Recovery tiers -- auto-degrades on rate-limit, never burns quota overnight.

Three Ways to Get Started

Starter
199 CNY99 CNY / one-time
First 10 clients only -- then back to 199 CNY
  • 30-min video walkthrough (Windows or macOS)
  • CC input coordinate calibration
  • Task Scheduler / launchd setup
  • Your use-case cron-prompt template
  • 1-week email support
Book 99 CNY Starter
Professional
999 CNY499 CNY / one-time
First 10 clients only -- then back to 999 CNY
  • Everything in Starter
  • 4 weeks 1-on-1 WeChat group support
  • Your agent decision-tree design
  • Weekly runtime log diagnosis
  • 5 cron-prompt templates (dev.to / multi-repo / scrape / WeChat / knowledge)
  • Monthly check-ins x 4
Book 499 CNY Professional
Enterprise
4999 CNY2499 CNY / one-time
First 10 clients only -- then back to 4999 CNY
  • Everything in Professional
  • 4 x 60-min video sessions (weekly)
  • Full orchestration customization for your team
  • 3-5 person team onboard (separate setups)
  • Team agent collaboration playbook
  • 30-day architecture guarantee
Book 2499 CNY Enterprise

All prices in CNY. Pay via Alipay / Wise / PayPal. No contract. Monthly retainer available after onboarding.

What the Setup Looks Like

Day 0 -- Contact plus Book

Email me with your setup (Windows/macOS) plus use case. I send you a time or WeChat handle.

Day 1 -- 30-Min Session

Screen-share walkthrough. Calibrate coordinates, set up Task Scheduler/launchd, run your first test fire. You will see it produce output within the session.

Day 2-7 -- Stabilize

Your agent runs overnight. You see hourly commits in your GitHub. I review loop.log and send tweaks if needed.

Day 8-plus -- Ship

Your agent handles backlog autonomously. You review and approve decisions. Time-to-market collapses. Pro clients get weekly log reviews for 4 weeks.

FAQ

I have zero coding experience -- can I do this?

Starter gets you set up even without coding experience -- as long as you can install Python venv and edit a markdown file. If you cannot articulate "what I want the agent to do" in writing, go Professional.

What is the difference vs. Anthropic API plus n8n?

API plus workflow platforms are for teams (1000-plus USD/month). This is for solo indies (20 USD/month Pro subscription). Run both in parallel: API for high-priority tasks, desktop loop for low-priority overnight work.

Will this get my account banned?

Desktop-loop simulates your own keyboard and mouse into your own IDE -- not a violation of Anthropic TOS. I have run 36-plus hours with zero warnings. Rules: do not run 24/7 at full speed, do not share the account, do not market it as "unlimited automation."

macOS or Windows -- which works better?

Both work. Windows Task Scheduler has friendlier GUI; macOS launchd is cleaner but requires a one-time Accessibility permission. Pick whichever machine can stay on 24/7.

What is the refund policy?

Starter is non-refundable (99 CNY = 30 min of my time). Professional and Enterprise: if after 2 weeks it is genuinely not working, partial refund considered case-by-case.

Can I resell this service to my clients?

Starter/Professional: no. Enterprise includes a commercial license for internal deployment, but you cannot offer "Claude Code 24/7 consulting" as a service to others.

Get the free Claude Code Readiness Checklist first

6 questions to answer before the call -- tells you if $99 Starter, $499 Professional, or DIY is right for your setup. Sent within 24 hrs, no spam.

GitHub is free. This service is not.

The full toolkit is MIT open source at github.com/jiejuefuyou/autoapp-toolkit -- read it, fork it, build it yourself.
What you are paying for is the 14 days of debugging I already did so you do not have to.

Email jiejuefuyou@gmail.com Browse Free GitHub