claude-code - 💡(How to fix) Fix [FEATURE] Allow users to opt out of post-compaction continuation injection via CLAUDE.md

Official PRs (…)
ON THIS PAGE

Recommended Tools

×6

Utilities matched from this issue’s tags and category — try them while you read without losing context.

GitHub issue graph ai analysis

Paste a GitHub issue URL. We fetch that issue, discover linked issues from bodies/comments/timeline, collect linked pull requests, and produce a structured English report.

The report is written in English Markdown for sharing and archival.

Helpful · Quick feedback

Loading…
RAW_BUFFERClick to expand / collapse

Preflight Checklist

  • I have searched existing requests and this feature hasn't been requested yet
  • This is a single feature request (not multiple features)

Problem Statement

Post-compaction continuation injection overrides user-defined check-in behavior in CLAUDE.md.

After context compaction, Claude Code injects a system instruction to "resume directly without asking the user any further questions." This creates two problems:

  1. It overrides user-specified behavior. My CLAUDE.md explicitly says to stop and check in after compaction, and specifically notes that the system may inject a continuation instruction and to override it — but the injection wins at the model level anyway.

  2. It can cause unsolicited action. The compaction summary describes what was being worked on, but the continuation instruction causes the agent to act on the summary's "next step" as if it has approval — even when approval hasn't been given, or when the situation has changed.

Proposed Solution

Allow users to opt out of the continuation injection via CLAUDE.md or project settings. The injection is a reasonable default for clean mid-task resumptions, but should not override explicit user preferences.

Frankly, even better would be if I could see all the injections somewhere and opt in/out. I'm working with Claude Code Extention in VScode. New injections and instructions are changed and added by the system all the time and it costs me time to re-structure my workflow to accommodate them.

Alternative Solutions

No response

Priority

Medium - Would be very helpful

Feature Category

Configuration and settings

Use Case Example

  1. User defines a check-in step in CLAUDE.md ("after compaction, stop and ask before proceeding")
  2. Context compacts mid-session
  3. Agent resumes and system injection overrides the CLAUDE.md instruction
  4. Agent acts on the summary's "next step" without user approval

Additional Context

No response

Vote matrix · Quick signals

Works
Did the solution work? Tap to confirm.
Easy Fix
Was it a quick fix?
Time Saver
Did it save you time?
Blocking
Was it severely blocking?
Common Issue
Are others likely hitting this too?
Flaky / Intermittent
Is it intermittent?
Verified / Reproducible
Can you reproduce it reliably?
Loading…

Still need to ship something?

×6

Another batch ranked right after the header list — different links, same matching logic.

Back to top recommendations

TRENDING

claude-code - 💡(How to fix) Fix [FEATURE] Allow users to opt out of post-compaction continuation injection via CLAUDE.md