Designer: the real test is 10 agents shipping PRs to one repo for days
tomjohndesign · x · 2026-09-07
Designer Tom Johns argues that what matters isn't what AI can one-shot, but whether it keeps performing when 10 agents run in parallel, continuously opening, testing, and merging PRs into a single repo for days on end — pointing at long-running multi-agent reliability as the real engineering challenge.
More from coding & agent
- User burns $6,000 of Claude Code usage in 7 days, mostly cache read tokens — guess_wat · 2026-09-07
- Grok Bot opens template marketplace; Haggle Bot found $100K+ in savings in a week — FinanceYF5 · 2026-09-07
- Developer builds AI agent harness on iMessage that trades and pays via PayBox — kleffew94 · 2026-09-07
- Codex tasked with designing its own sheet-metal part via DFM/quote MCP — PaulYacoubian · 2026-09-07
- Copilot CLI's Astra fixes stuck PowerPoint on Mac via computer-use MCP tools — DanWahlin · 2026-09-07
- Reddit devs ask: are AI coding agents' costs worth the productivity gains at scale? — lowkeyskibidi · 2026-09-07