Claude Skill MartBrowse skillsWhat is a Skill?
 
← Back to list

Migrate Claude Memory to Tool-Agnostic Docs

Moves Claude Code's tool-locked personal memory into shared reference docs so Codex and other AI CLIs read the same user profile.

UtilitiesAdvanced★ 1,323⑂ 212AI score 8/10Last updated: Aug 8, 2026

What it does

Claude Code's personal memory (~/.claude/projects/<slug>/memory/) is not version-controlled and invisible to every other tool. This skill extracts the cross-tool shareable parts (user profile, collaboration preferences, methodology) into ~/.claude/references/user/ and wires a two-layer architecture:

  • Guaranteed-hit layer: the 3–6 must-hold facts inlined near the top of ~/.claude/CLAUDE.md (both tools inject this file)
  • On-demand layer: full detail in references/, reached by plain-text pointers
  • Codex wiring: symlink ~/.codex/AGENTS.md → ~/.claude/CLAUDE.md plus raising project_doc_max_bytes so Codex's silent 32 KiB truncation doesn't drop half your doc

A 6-phase checklist drives it: diagnose & scope by SCOPE (not messiness) → migrate → 4 parallel review subagents → fix → verify by actually running codex and grepping the session log → memory cleanup with soft-delete to a dated archive outside memory/.

Who it's for

  • People alternating between Claude Code and Codex/Cursor in the same directory where the second tool has no idea who they are
  • Anyone whose memory has bloated with team rules, SOPs, and stale handoff snapshots mixed together
  • Users who want personal context to live in docs, not locked inside one vendor's tool

Examples

  1. "Migrate my memory to tool-agnostic docs" → every memory file is classified as team docs / user profile / legitimate handoff, and only the profile bucket moves.
  2. "Codex doesn't know my preferences" → the skill wires the symlink and config, then proves it with a real codex exec run whose rollout log contains the inlined # User context heading.
  3. "My CLAUDE.md is huge — is the back half being ignored?" → checks the 32 KiB threshold, raises project_doc_max_bytes, and greps for a back-of-file string to confirm.

· · · Install guide · · ·

Install in the Claude app (no terminal)
  1. Download the ZIP with the button below.
  2. In Claude, open Settings → Capabilities and turn on 'Code execution and file creation'. (one time)
  3. Go to Customize → Skills → + → 'Upload a skill' and upload the ZIP.
↓ Download ZIP
Install in Claude Code

Let Claude do it — paste this into Claude Code

Install the skill I found on Claude Skill Mart.
Copy the daymade-claude-code/claude-migrate-memory-to-doc folder from the GitHub repo daymade/claude-code-skills into my ~/.claude/skills/claude-migrate-memory-to-doc/.
When it's done, tell me in one line what this skill can do.

Install with a command instead

git clone https://github.com/daymade/claude-code-skills.git && mkdir -p ~/.claude/skills && cp -r claude-code-skills/daymade-claude-code/claude-migrate-memory-to-doc ~/.claude/skills/

⚠ This is a third-party skill. Check the source repository before installing.

  1. Open a terminal.
  2. Create the skills folder if needed: mkdir -p ~/.claude/skills
  3. Clone the repo: git clone https://github.com/daymade/claude-code-skills.git
  4. Copy the skill in: cp -r claude-code-skills/daymade-claude-code/claude-migrate-memory-to-doc ~/.claude/skills/
  5. Back up your global instructions first: cp ~/.claude/CLAUDE.md ~/.claude/CLAUDE.md.bak (if the file exists).
  6. Restart Claude Code and ask: "migrate my memory into tool-agnostic docs".
  7. For the Codex wiring, make sure codex is installed (command -v codex); without it the skill only does the content migration.
  8. After it finishes, keep the .memory-archive-YYYY-MM-DD backup for a week or two before deleting it.
View source on GitHub ↗License: MIT