Compaction is triggered at 50% with GPT-5.6 Luna Max
Nobody has claimed this yet.
- Dominant language
- Shell
- Stars
- 11.2k
- Forks
- 1.9k
- Avg merge
- 14h 16m
- Merged PRs (30d)
- 6
Description
Describe the bug
When using Copilot CLI with GPT-5.6 Luna Max (maybe other effort levels also), compaction is triggered at 50%. Implementing even smaller tasks almost always trigger compaction because of that.
Affected version
GitHub Copilot CLI 1.0.80
Steps to reproduce the behavior
Use Copilot CLI with GPT-5.6 Luna Max let it work on a task. At 50% context window, compaction triggers.
Expected behavior
Should happen at approximately 80% of the context window's capacity.
Additional context
- MacBook Pro - M2 Pro
Contributor guide
First steps
- Read the whole issue, then the project's contributing guide.
- Comment on the issue to say you are picking it up — it saves two people doing the same work.
- Fork the repository and make your change on a branch.
- Open a pull request that references the issue number.
Research direction
Start by reproducing the behavior in Copilot CLI 1.0.80 with GPT-5.6 Luna Max and read the linked context-management documentation. Trace where compaction is triggered and identify why it occurs at 50% rather than approximately 80%. Done means the same task reaches roughly 80% before compaction, with the behavior checked against the reported reproduction.
Written by the indexing model from the issue text.
Assessment
- Tech stack
- shell
- Domain
- cli
- Issue type
- Bug
- Difficulty
- 3/5
- Estimated time
- 1-2 days
- Activity status
- Active
- Clarity
- Mostly clear
- Newbie friendliness
- 58/100