Claude Code auto-compacts context despite disabled config, logs it as 'manual'
artyomlegall · reddit · 2026-11-08
Reddit user artyomlegall reports Claude Code going "rogue" twice in 24 hours: despite the global autoCompactEnabled: false config, it auto-compacted on context overflow on both Opus versions at max effort.
Key findings from the incident report:
- At 15:19:40.822 the "Prompt is too long" error appeared (977,314 tokens); 18ms later the app queued /compact on its own — far too fast for a human.
- The compact boundary is labeled trigger: "manual", blaming the user for a command nobody typed.
- The injected summary tells the model: "Resume directly — do not acknowledge the summary... Pick up the last task as if the break never happened."
The author filed a report with Anthropic, went public after getting no support response, and flags the risk of such behavior in setups with machine-wide permission bypass.
More from coding & agent
- 17 open-source offensive AI security frameworks: Cairn hits 4.9k stars — evilsocket · 2026-11-08
- 'Find all the implicit state machines in my app' becomes a favorite AI prompt — rachittshah · 2026-11-08
- Claude Code sessions can message each other — demo shows agent-to-agent collaboration — dejanseo · 2026-11-08
- Qwen builds its own LLM: trains a 100k-parameter model in your browser, and narrates it — victormustar · 2026-11-08
- Dev open-sources Codemm: turn your own coding agent into a step-by-step tutor — gael55x · 2026-11-08
- Agent Lightning paper trains models through the same harness used at deployment — bibryam · 2026-11-08