Codex Terra and Luna sub-agents stop doing work around 200K-context runs
runfence · ghdev · 2026-07-18
A Codex user reports that the Terra and Luna sub-agents sometimes stop their turn without doing anything once the context grows to around 200k tokens.
- The issue was observed during a review task asking the agent to score the usefulness of added tests.
- The user then repeatedly asked follow-up questions like whether anything was missed or whether all bodies and code paths were read.
- Instead of continuing, the agent would stop after about 30 seconds.
- The reporter says the same pattern also showed up repeatedly in a larger 400 KB review-report workflow.
More from coding & agent
- Spring AI 2.1.0-RC1 adds tool-call confirmation, pausing agent loops before real-data access — therealdanvega · 2026-10-26
- 250 Builders Hack With Codex at Berlin Community Hackathon Ahead of OpenAI DevDay — paw_lean · 2026-10-26
- Building AI bots on raw APIs? Devs say models are 'dumb without the harness' — madeofAjala · 2026-10-26
- HarnessMM cuts agent tokens per task by up to 88% while raising success rates — omarsar0 · 2026-10-26
- Microsoft demos intent-classifying agents built with Copilot App and Foundry, deployable across Teams and Outlook — marlene_zw · 2026-10-26
- Looking for a wearable mic with best battery life that feeds recordings to personal agents via MCP — altryne · 2026-10-26