Memory Transfer Skill for Openclaw

A specialized utility for moving and transforming memory files between different AI agent instances to maintain context across workflows.

codeblackhole1024
v2.1.0
Mar 7, 2026
0
788
0

Install & Download

1. ClawHub CLI

The fastest way to install a skill directly from the registry.

npx clawhub@latest install memory-transfer-enhanced

2. Manual Installation

Copy the skill folder to one of these locations

Global
~/.openclaw/skills/
Workspace
<project>/skills/

Priority: Workspace > Local > Bundled

3. Prompt Installation

Copy this prompt to OpenClaw to install it automatically.

Help me install memory-transfer-enhanced using Clawhub. If Clawhub is not installed, install it first (npm i -g clawhub).

Prefer to download?

Get the raw skill files in a ZIP archive.

What is Memory Transfer Skill?

The Memory Transfer skill is a robust tool designed to synchronize knowledge and context across your ecosystem of Openclaw Skills. It allows developers to migrate long-term memories, daily logs, and topic-specific data from one agent to another, ensuring that specialized sub-agents have the background information they need to succeed. By managing the flow of information, this skill prevents context fragmentation and reduces the need for repetitive prompting.

Whether you are spinning up a new assistant or backing up an existing one, this addition to your Openclaw Skills library provides the technical infrastructure to move data safely. It handles the nuances of agent identity, ensuring that when knowledge is shared, the target agent understands the perspective from which the information originated, rather than inheriting a confused self-identity.

Memory Transfer Skill Use Cases

  • Migrating historical context from a primary general-purpose agent to a specialized sub-agent.
  • Sharing specific technical topic memories between multiple Openclaw Skills to keep them synchronized.
  • Creating verbatim backups of agent memory files before performing system resets or version updates.
  • Cleaning personal user data from memory files while transferring pure knowledge to a new agent instance.

How Memory Transfer Skill Works

  1. Identify the source and target agent IDs by listing available directories in the OpenClaw configuration path.
  2. Choose between share mode for knowledge transfer or clone mode for exact data duplication.
  3. Optionally define a topic or keyword to filter which specific memory files should be included in the transfer.
  4. Run a dry-run execution to preview how the content will be transformed and which files will be affected.
  5. Confirm the operation to move the data, applying identity adaptation rules to the Markdown content during the process.

Memory Transfer Skill Setup

To begin using this tool with your Openclaw Skills, locate your agent workspace IDs first:

ls /home/node/.openclaw/

Perform a basic transfer between two agents using the following command:

node memory-transfer.js transfer <source-agent-id> <target-agent-id>

For a guided experience, use the interactive mode which prompts for topics and safety confirmations:

node memory-transfer.js transfer <source> <target> --mode interactive

Memory Transfer Skill Data Schema & Taxonomy

This skill interacts with the standard storage structure used by Openclaw Skills to ensure compatibility across different agent types:

File Type Location Description
Long-term Memory MEMORY.md The primary file containing consolidated agent knowledge.
Daily Logs memory/YYYY-MM-DD.md Date-stamped files capturing specific daily interactions.
Topic Files memory/*.md Specialized files dedicated to specific keywords or projects.

During share mode, the skill applies a transformation schema that maps source pronouns (e.g., 'I', 'my') to target perspectives (e.g., 'Agent Main', 'your') to maintain consistent identity.

Memory Transfer Skill Advanced Features

  • Topic-specific filtering using the --topic flag to isolate and move only relevant knowledge clusters.
  • Identity Adaptation Engine that intelligently rewrites role statements so the target agent doesn't inherit the source agent's persona.
  • Privacy Filtering which automatically strips personal information like phone numbers, addresses, and emails during sharing operations.
  • Dry-run validation prevents accidental overwrites by showing a full diff of changes before they are committed to the file system.
  • Search and List functions to audit the contents of different Openclaw Skills before initiating a migration.

SKILL.md


Loading

Related Openclaw Skills

METADATA

Github Stars: 0
forks: 0

Featured*