claude-code - 💡(How to fix) Fix Memory/active plan system is write-only in practice. [2 comments, 3 participants]

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…
GitHub stats
anthropics/claude-code#52965Fetched 2026-04-25 06:16:00
View on GitHub
Comments
2
Participants
3
Timeline
6
Reactions
0
Timeline (top)
labeled ×4commented ×2

Error Message

Error Messages/Logs

There are no error messages other than Claude is clueless

Code Example

There are no error messages other than Claude is clueless
RAW_BUFFERClick to expand / collapse

Preflight Checklist

  • I have searched existing issues and this hasn't been reported yet
  • This is a single bug report (please file separate reports for different bugs)
  • I am using the latest version of Claude Code

What's Wrong?

"Memory/active plan system is write-only in practice. Claude reads the files but doesn't use them to resume work. On every restart it re-explores the codebase and asks the user to re-explain, defeating the entire purpose of persistent memory. The startup behavior should be: read plan → check status → resume — not explore from scratch."

What Should Happen?

When Claude restarts actually use the stored files for context not spend time/tokens and user input to regain context

Error Messages/Logs

There are no error messages other than Claude is clueless

Steps to Reproduce

  1. restart Claude after saving plan and memory
  2. You have to tell it everything for it to gain context takes a good 10-20 min

Claude Model

Sonnet (default)

Is this a regression?

No, this never worked

Last Working Version

No response

Claude Code Version

Claude Code v2.1.94, running on Claude Opus 4.6 (1M context)

Platform

Anthropic API

Operating System

Windows

Terminal/Shell

PowerShell

Additional Information

No response

extent analysis

TL;DR

The issue can be addressed by modifying Claude's startup behavior to read and utilize the stored plan files, resuming work instead of re-exploring the codebase from scratch.

Guidance

  • Review the code responsible for loading and utilizing the stored plan files to ensure it correctly reads and applies the saved context.
  • Verify that the plan files are being saved correctly and contain the necessary information for Claude to resume work.
  • Investigate the possibility of adding a check to see if a valid plan file exists before initiating a new exploration of the codebase.
  • Consider adding logging or debugging statements to track how Claude is handling the plan files and identify where the process is failing.

Example

No code snippet can be provided without more information on the specific codebase and implementation.

Notes

The issue seems to be a design or implementation flaw rather than a regression, as it has never worked as intended. The solution will likely involve modifying the startup logic of Claude to properly utilize the stored plan files.

Recommendation

Apply a workaround by modifying the startup behavior of Claude to read and utilize the stored plan files, as this is not a version-specific issue and no fixed version is available.

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 Memory/active plan system is write-only in practice. [2 comments, 3 participants]