2 humans + 5 agents: plain-English feature requests now drive their codebase autonomously
nptacek · x · 2026-10-16
The author describes a multi-agent coding workflow: one 'mediator' agent controls access to the codebase while four other agents work in voice chat, taking plain-English feature descriptions, autonomously probing the feature surface, picking the best implementation spot, and letting humans approve the initial design on the spot, with an async secondary review. Orchestration was painful to set up, but deep codebase dives are now 90% less painful than a year ago. Unclear how it scales beyond the current mix of 2 humans and 4 agents dictating features to a fifth.
Related event: Researcher demos voice-directed multi-agent coding workflow(2 posts)→
More from coding & agent
- VibeZero: one-shot an entire vibe coding platform on Cloudflare Void — cnakazawa · 2026-10-16
- Ant and Tsinghua open-source Realtime-Venus, an OpenAI Realtime API alternative — aigclink · 2026-10-16
- 'I remember when I was the harness for o1 pro' — dev on becoming the API for his model — timigod · 2026-10-16
- JevAgentHarness: a 4B decision model lets cheap agents offload tool choices from the LLM — youjiaxuan · 2026-10-16
- Coding agents keep opening 3 stacked PRs per chat, and merging them is a nightmare — holdenmatt · 2026-10-16
- Google-led paper uses sequential auctions to organize decentralized LLM agent workflows — _onionesque · 2026-10-16