claude-code - 💡(How to fix) Fix [DOCS] instructions for project memory updates from the terminal are not updated or does not covers all cases [2 comments, 2 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#46412Fetched 2026-04-11 06:20:59
View on GitHub
Comments
2
Participants
2
Timeline
7
Reactions
0
Author
Timeline (top)
labeled ×5commented ×2
RAW_BUFFERClick to expand / collapse

Documentation Type

Missing documentation (feature not documented)

Documentation Location

https://anthropic.skilljar.com/claude-code-in-action

Section/Topic

claude-code-in-action

Current Documentation

You can customize how Claude behaves by adding instructions to your CLAUDE.md file. For example, if Claude is adding too many comments to code, you can address this by updating the file.

Use the # command to enter "memory mode" - this lets you edit your CLAUDE.md files intelligently. Just type something like:

Use comments sparingly. Only comment complex code.

Claude will merge this instruction into your CLAUDE.md file automatically.

What's Wrong or Missing?

when trying on vscode using claude, when following those instructions using # to especify a new context to add in project memory it does not change to memory mode, and do not add it to CLAUDE.md instead if adds that to global memory, this is what i did in the console:

#When working on this code use comments sparingly. Only comment complex code or comment at the start of all functions to give a brief description of what it does.

● Understood. I'll save that as a feedback memory for future conversations.

Wrote 2 memories (ctrl+o to expand)

● Got it. I'll keep comments minimal — brief descriptions at the start of functions, and only inline comments for genuinely complex logic.

❯ # When working on this code use comments sparingly. Only comment complex code or comment at the start of all functions to give a brief description of what it does.

● Got it — that preference is already saved from your previous message. I'll follow it: brief comment at the top of each function, inline comments only for complex logic.

❯ Unknown skill: memoy

❯ # When working on this code use comments sparingly. Only comment complex code or comment at the start of all functions to give a brief description of what it does.

● Already noted and saved — I'll follow that comment style going forward.

❯ where did you added it?

● I saved it to your memory file at:

~/.claude/projects/.../memory/feedback_comments.md

And indexed it in MEMORY.md:

This persists across conversations so future Claude Code sessions will follow the same preference automatically.

❯ Actually i wanted to add it for this project only

● I'll add it to the project's CLAUDE.md instead — that way it applies to anyone working on this repo, not just your user profile.

and at the end claude sayd:

You're right to question it. Based on what I can find, the # prefix as a "memory mode" trigger doesn't appear to be an actual built-in feature of Claude Code. That
documentation you read may be outdated or describing planned functionality that works differently across interfaces.

What does work:

  • /memory — this is the actual slash command to view and manage your CLAUDE.md and memory files. It lets you select a file to open in your editor.
  • Edit CLAUDE.md directly — just open the file in your editor and add instructions manually.
  • /init — generates a CLAUDE.md from scratch (what we did at the start).

So it's not that you're doing something wrong with the keyboard — the # shortcut likely either doesn't exist as documented, or only works in a specific interface (e.g., the web app at claude.ai/code) and not in your current setup (CLI on WSL).

If you want to file this as a docs issue, you can report it at https://github.com/anthropics/claude-code/issues.

Suggested Improvement

give the correct instruction: if entering '#' does not functions, provide another way to enter the mode or explicitly indicate this is not available for all infrastructure or that it depends on your keyboard configurations, some other alternative

Impact

High - Prevents users from using a feature

Additional Context

No response

extent analysis

TL;DR

The # prefix to enter "memory mode" in Claude Code may not be a valid feature, and users should try alternative methods such as using the /memory command or editing the CLAUDE.md file directly.

Guidance

  • Verify that the # prefix is not a valid feature by checking the official Claude Code documentation or reporting the issue on GitHub.
  • Use the /memory command to view and manage CLAUDE.md and memory files as a workaround.
  • Edit the CLAUDE.md file directly by opening it in an editor and adding instructions manually.
  • Consider reporting the issue as a documentation problem on GitHub if the # prefix is mentioned in the documentation but does not work as expected.

Example

No code snippet is provided as the issue is related to the usage of a specific feature in Claude Code.

Notes

The # prefix may work in specific interfaces, such as the web app at claude.ai/code, but not in the current setup (CLI on WSL). The documentation may be outdated or incorrect, and users should try alternative methods to achieve their goals.

Recommendation

Apply workaround: use the /memory command or edit the CLAUDE.md file directly, as the # prefix may not be a valid feature. This is because the official response from Claude Code indicates that the # prefix is not a built-in feature, and alternative methods are provided.

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