# Remote Coding Agents: Ship Verified Pull Requests Anywhere
Run Your AI Coding Team From Your Phone and Ship More in 30 Days

- Price: $12.99 (Kindle ebook)
- Level: Advanced
- Pages: 265
- Best for: Anyone leaving the laptop half open so the agent keeps running
- You finish with: A phone-to-PR system that returns tested pull requests from isolated workspaces, with a cost per accepted change and a hard monthly cap
- Series: The Overnight AI Team, book 3 of 3
- Buy on Amazon: https://amzn.to/4xvtfh0
- Web page: https://youcanbuildthings.com/books/remote-coding-agents

## What you get

Stop leaving the laptop half open so the agent keeps running. Send a job from your phone and get back a pull request with tests, a build receipt, and evidence you review in sixty seconds. Every job runs in its own workspace, reports to one board, and stops at a cost ceiling you set.

Developers are sending real repository work away from the desk and getting finished branches back. The difference isn't more tokens. It's a remote operating system that keeps every job visible, bounded, and reviewable.

This book builds that system: a job contract an agent can execute, a phone-to-PR route that survives a dropped connection, an isolated workspace for every job, a gate that blocks weak pull requests automatically, and a budget tied to accepted changes instead of tokens.

The free companion repo ships the complete remote-agent scaffold, runnable offline with no account, key, or network connection.

## Chapters

1. The Pull Request Waiting When You Get Back: The three receipts that separate a remote system from a remote trigger: a refusal, a returned change, and a priced decision.
2. Write the Job Before You Send It Away: Write a dispatch contract with permissions and stop conditions, and watch the validator refuse a bad job.
3. Pick the Runtime With Numbers, Not Loyalty: Run a paired trial of two runtimes on your own repo and write the runtime rules your measurements support.
4. Send One Safe Change From Your Phone: Dispatch one change from your phone and get back its own branch with a commit-bound test and build receipt.
5. One Board for Every Session: Build one status board for every live session, with leases, heartbeats, and a reconnect that survives a dropped connection.
6. Give the Agent a Room, Not the Building: Give every job an isolated workspace and a fail-closed policy, then capture the refusal when it reaches too far.
7. No Green Check, No Pull Request: Gate every pull request on tests and builds with a check the authoring run cannot approve or merge.
8. Make the Handoff Readable in Sixty Seconds: Attach a hashed evidence bundle and a generated handoff you can review cold in sixty seconds.
9. Price an Accepted Change, Not a Token: Price cost per accepted change from a five-job ledger, find break-even, and set a hard monthly cap.
10. Add Capacity Without Adding a Manager: Add workers with transactional locks and a global stop, and find the highest concurrency that still pays.
11. Break It Before It Breaks You: Kill jobs on purpose, restore from backup, and replay the same job from a tested incident runbook.
12. Your Next Thirty Days of Verified Shipping: Drop the scaffold into a fresh repo and roll it out through four go/no-go gates over thirty days.

## Free guides from this book

- [How to Keep Claude Code Working After You Close Your Laptop](https://youcanbuildthings.com/articles/run-coding-agents-remotely.md): Send a job from your phone, close the laptop, and come back to a finished branch by running Claude Code somewhere other than the laptop. Remote Control runs on your machine and GitHub's cloud agent stops at 59 minutes, so here's how to pick the route and a 10-minute test that proves it works.
- [How to Merge AI Agent PRs From Your Phone With GitHub Actions](https://youcanbuildthings.com/articles/run-ai-coding-agents-in-ci-cd.md): Merge agent-written pull requests from your phone knowing every required check actually ran on that commit. GitHub counts a skipped job as a success, so one job that always runs checks each result by name, and it failed the skipped type check on the first test.
- [How to Review a Coding Agent's PR in 60 Seconds](https://youcanbuildthings.com/articles/review-ai-generated-pull-requests.md): Approve or reject a coding agent's pull request in about a minute, from your phone, because the screen is built from an evidence file and not from the agent's summary. The same script refuses to show anything if the branch moved or the file was edited after the run.
- [How to Run Multiple Coding Agents at Once Without Babysitting](https://youcanbuildthings.com/articles/run-multiple-coding-agents-at-the-same-time.md): Run ten coding agents at once and see in one query which one needs you, even after the terminal that started it is gone. The board is one SQLite table, and it catches the worker that is still checking in but hasn't done anything for 18 minutes.
- [What Claude Code's Skip Permissions Mode Does, and How to Use It](https://youcanbuildthings.com/articles/claude-code-dangerously-skip-permissions.md): Claude Code's --dangerously-skip-permissions flag turns off every permission prompt, so a job runs start to finish while you're away from the desk. Here's what it stops checking, and the four settings that keep it out of your CI files and everything outside the project.