宝玉 on doc-first multi-agent coding: write a spec, review, then let Opus execute
ciaotigre · x · 2026-09-22
宝玉 shares his doc-first AI coding workflow instead of plan mode: a human reviews the spec for direction, Fable writes the doc with technical details and validation steps, Opus executes, then Fable verifies against the doc in a fresh session. He also sometimes uses GPT-6 Pro in ChatGPT to write specs from the repo, mixes Claude Code with Codex, and keeps key docs updated. His point: the value lies not in plan mode but in a human-confirmable doc that carries context across multiple agents.
Related event: Developers Favor Spec Documents Over Plan Mode for AI Coding(4 posts)→
More from coding & agent
- User automates entire video clipping workflow with Nous Research's Hermes agent — Teknium · 2026-11-12
- As models improve, writing good evals gets harder and more important — amankhan · 2026-11-12
- Prime agent ships with day-one multi-agent communication, runs in Minecraft unmodified — sarahcat21 · 2026-11-12
- Autonomous agent edits just 1,024 weights to lift Qwen3.5 0.8B from 7.4% to 37.4% on MATH-500 — VictorSanh · 2026-11-12
- Omarchy 4.6 lets you share local AI models across your tailnet, says DHH — alexcovo_eth · 2026-11-12
- Coding harnesses hand shell access to inference providers — 4-step gateway mitigation — shensi · 2026-11-12