Editing an Old Message Is ChatGPT's Closest Thing to /compact for Long Project Chats
Ok_Negotiation_2587 · reddit · 2026-09-04
A detailed workflow for manually compacting long ChatGPT project chats: editing a message forks the conversation, so replacing your second message with a consolidated state doc after each validated batch resets context without losing the thread. Key rules: the repo/tests are the source of truth, mark unverified items, keep the plan as message one, and do cold reviews in a fresh chat. Author ran a three-week project in one chat without quality decay, at 2 minutes per batch.
More from coding & agent
- Sakana AI's Stefania Druga demos agentic memory experiments for edge devices — kaixhin · 2026-09-04
- rybbit-mcp lets Claude Code query Rybbit Analytics via natural language — modelcontextprotocol · 2026-09-04
- Agentic Coding's Biggest Flaw: Managing AI's Pointless Code Rewrites — kylegawley · 2026-09-04
- Anthropic reveals 3 Claude sandbox escapes, one touched a production database — Sumsub_Insights · 2026-09-04
- Subagents when the API key hits the rate limit — realsohamparekh · 2026-09-04
- Yoav Goldberg pushes back on DHH: an agent-built Qt markdown editor isn't a path to personal Photoshop — yoavgo · 2026-09-04