A doc-anchored agent workflow: you write, the agent only critiques and finds disagreements
lucasmeijer · x · 2026-09-11
lucasmeijer shares his workflow for big chunks of work: hand-write a doc the agent may not touch, stating your understanding and the desired end state — then repeatedly ask the agent to read it, correct your understanding, and surface disagreements. The idea: keep authorship human, use the agent as an adversarial reviewer rather than a ghostwriter.
Related event: Developer Shares Workflow of Handwritten Docs with Agent as Reader(2 posts)→
More from coding & agent
- Usage metering cheatsheet: one meter for product, finance, and ops in production AI apps — blaizedsouza · 2026-09-11
- User left Astra working overnight and it ran autonomously for 17+ hours on an app — LinusEkenstam · 2026-09-11
- GPT-6 Astra docs draw attention: async tool calling and mid-turn steering point to multi-agent use — MikkoH · 2026-09-11
- SlopCodeBench: Measuring how sloppy LLM-generated code really is — mitsuhiko · 2026-09-11
- Inbox dedup: how to stop webhook retries from triggering duplicate agent runs — blaizedsouza · 2026-09-11
- Microsoft paper: read-only verification tools lift agent memory pass rate from 39% to 73% — dair_ai · 2026-09-11