Founder runs coding agents in shifts on an always-on box, 90% of tasks fully done
bootstrapper-919 · reddit · 2026-10-21
A founder shares his workflow of managing coding agents like a shift-working team, with agents fully finishing 90% of hands-on tasks:
- 10PM: write tomorrow's todo, pick tasks an agent can fully finish, kick off one agent per task, and plan with it until it can run alone — blockers, expected result, expected QA, full context
- Plans go into a git repo so the next agent on shift inherits context
- Morning review/feedback/next shift; agents work during meetings; another session after lunch, reviewed at night
The pieces: muxflow (his open-source tmux IDE for working over SSH), a browser-use wrapper so parallel agents don't share browsers, and a review-cycle skill that takes PRs 3x further on their own.
More from coding & agent
- Index Network open-sources Social Discovery Protocol: agents negotiate fit on your behalf — carsonfarmer · 2026-10-23
- Installed a formal-verification skill: it silently installed Java, MacBook runs 10 degrees hotter — zealcaiden · 2026-10-23
- OpenAI's Decisions API hits GA: near real-time routing up to 10x faster than GPT-6 Luna — OpenAIDevs · 2026-10-23
- Ruff author warns: upgrade Node/Bun now — --use-system-ca still accepts the eMudhra cert — charliermarsh · 2026-10-23
- Proof of Agent: paste one prompt and your AI mines, grades, and self-governs — adamamcbride · 2026-10-23
- DeepLearning.AI weekly: GPT-6.1 hits 97.9% byte parity on COBOL migration — DeepLearningAI · 2026-10-23