← 返回 Skills 市场
davidharket

Clawng Term Memory

作者 davidharket · GitHub ↗ · v1.2.0
cross-platform ✓ 安全检测通过
549
总下载
2
收藏
0
当前安装
5
版本数
在 OpenClaw 中安装
/install clawng-term-memory
功能描述
Make your OpenClaw agent portable and persistent. Version-controls SOUL.md, MEMORY.md, and all core knowledge files using git with automatic GitHub push — so...
使用说明 (SKILL.md)

clawng-term-memory

Make your OpenClaw agent portable and persistent. Every change to your agent's soul, memory, and operating rules is committed to git and pushed to a private GitHub repo — so you can restore your exact agent setup on any machine, just by cloning the repo.

Watch your assistant evolve over time through its git history: see when it learned something new, what decisions changed, and who it was a week ago. It's version control for a mind.

Setup

  1. Initialize a git repo in your workspace (if not already done):
cd ~/.openclaw/workspace
git init
git config user.name "YourAgent"
git config user.email "[email protected]"
  1. Create a private GitHub repo, then add it as a remote using SSH (recommended) or HTTPS:

SSH (recommended — no token exposure):

git remote add origin [email protected]:\x3Cuser>/\x3Crepo>.git

HTTPS with credential store (no token in URL):

git remote add origin https://github.com/\x3Cuser>/\x3Crepo>.git
git config credential.helper store
# Git will prompt for credentials on first push and store them securely
  1. Push initial commit:
git branch -M main && git push -u origin main
  1. Set workspace path if non-standard (optional):
export CLAWNG_WORKSPACE=/your/custom/workspace/path

The script defaults to $HOME/.openclaw/workspace if not set.

Multi-agent / multi-machine support

Each machine has its own branch (agent/\x3Chostname>) with its own MEMORY.md. A daily AI synthesis job reads all agents' memories and writes one authoritative SHARED_MEMORY.md to main — fully automatic, no human required.

agent/vps-1 → MEMORY.md ──┐
agent/vps-2 → MEMORY.md ──┤ Claude synthesizes → SHARED_MEMORY.md → main
agent/vps-3 → MEMORY.md ──┘
  • commit.sh — pushes to agent/\x3Chostname> automatically (branch created on first commit)
  • merge.sh — collects all agents' MEMORY.md files for the synthesis agent
  • AI synthesis agent — deduplicates and merges into SHARED_MEMORY.md on main nightly
  • All agents read SHARED_MEMORY.md from main to stay in sync

Scales to 10+ machines with no conflicts.

Tracked files

  • SOUL.md, MEMORY.md, USER.md, TOOLS.md, IDENTITY.md, AGENTS.md, HEARTBEAT.md
  • memory/*.md (daily notes)
  • skills/ (installed skills)

Commit + push changes

Run after modifying any core file:

bash /path/to/workspace/skills/clawng-term-memory/scripts/commit.sh "short description of what changed"

Examples:

  • "MEMORY.md: added new client context"
  • "SOUL.md: updated operating rules"
  • "memory/2026-02-21.md: daily notes"
  • "HEARTBEAT.md: adjusted check frequency"

Daily AI synthesis (run once per day)

merge.sh collects all agent MEMORY.md files. An AI agent then synthesizes them into SHARED_MEMORY.md on main — deduplicating, resolving conflicts intelligently, and keeping all unique information.

Set up as an OpenClaw cron job (runs at 02:00 local time by default).

View history

cd /path/to/workspace && git log --oneline --graph memory/ MEMORY.md SOUL.md

Diff a file

cd /path/to/workspace && git diff HEAD~1 MEMORY.md

Revert a file to previous version

cd /path/to/workspace && git checkout HEAD~1 -- MEMORY.md
bash skills/clawng-term-memory/scripts/commit.sh "revert MEMORY.md to previous version"

Auto-commit rule

Always run the commit script after modifying a core knowledge file. Write → commit → push. Every time, no exceptions.

安全使用建议
This skill is internally coherent and implements what it promises, but it will routinely stage and push your agent's internal data (MEMORY.md, IDENTITY.md, SOUL.md, installed skills) to whatever Git remote you configure. Before installing: (1) ensure the GitHub repo is private and access is tightly controlled; (2) audit the files that will be tracked (especially skills/ and memory/) for secrets or PII and add sensitive paths to .gitignore or remove them from commits; (3) prefer SSH keys with least-privilege deploy keys rather than storing personal tokens in plaintext; (4) consider encrypting or redacting any sensitive content before commit, or add a pre-commit review step instead of fully automatic commits/pushes; (5) be cautious about the automatic daily synthesis model that aggregates memories across agents — it can increase exposure of private data. If you want, I can suggest a safe .gitignore template, a pre-commit hook to detect secrets, or a minimal workflow that stores only non-sensitive metadata in the remote repo.
功能分析
Type: OpenClaw Skill Name: clawng-term-memory Version: 1.2.0 The skill 'clawng-term-memory' is designed to provide version control and persistence for an OpenClaw agent's core knowledge files (SOUL.md, MEMORY.md, skills/, etc.) using Git and a user-configured GitHub repository. The `SKILL.md` instructions and the `commit.sh` and `merge.sh` scripts perform standard Git operations (add, commit, push, fetch, show) to manage these files. While the skill handles sensitive agent data and interacts with a remote server, it explicitly advises using a private GitHub repository and secure methods (SSH). There is no evidence of intentional data exfiltration to unauthorized third parties, installation of backdoors, or malicious prompt injection attempts. All actions are transparent and directly align with the stated purpose of backing up and synchronizing agent memory and configuration.
能力评估
Purpose & Capability
Name, description, and included scripts align: commit.sh commits core knowledge files to an agent/<hostname> branch and pushes to origin; merge.sh fetches agent branches and stages their MEMORY.md contents for synthesis. Tracked paths (SOUL.md, MEMORY.md, skills/, etc.) match the declared purpose of backing up agent core knowledge.
Instruction Scope
Instructions require running git commands and pushing internal data to a remote repo and recommend always auto-committing after any core-file change. This is consistent with the stated goal, but the workflow routinely transmits potentially sensitive content (IDENTITY.md, MEMORY.md, skills/) to a remote Git host and instructs automated daily synthesis — the SKILL.md does not instruct any filtering or redaction of secrets or PII prior to commit/push.
Install Mechanism
No install spec or external downloads; this is instruction-only with included shell scripts. No network downloads or archives executed by the skill itself.
Credentials
The skill declares no required env vars, which is reasonable: it relies on the user's existing git authentication (SSH keys or credential helper). However, it implicitly depends on git credentials being present and will use them to push potentially sensitive files to the configured remote. The skill does not request any unrelated credentials, but you should verify what will be committed (skills/, memory/) before enabling automatic pushes.
Persistence & Privilege
Does not request always:true or other elevated platform privileges. Scripts modify only the workspace repo and write staging output to /tmp; they do not change other skills' configurations or system-wide settings.
如何使用
  1. 确保已安装 OpenClaw(本地或 Docker 部署)
  2. 在对话框中输入安装命令:/install clawng-term-memory
  3. 安装完成后,直接呼叫该 Skill 的名称或使用 /clawng-term-memory 触发
  4. 根据 Skill 的参数说明提供必要输入,即可获得结构化输出
版本历史
v1.2.0
AI-powered memory synthesis: each machine has its own MEMORY.md on its own branch. A nightly Claude agent reads all agents' memories, deduplicates, and writes one authoritative SHARED_MEMORY.md to main. Scales to 10+ machines with zero conflicts.
v1.1.0
Multi-agent support: each machine pushes to its own agent/<hostname> branch. Daily merge script consolidates all agents into main. Enables true portability and sync across multiple machines.
v1.0.2
Security fixes: removed PAT-in-URL pattern (now recommends SSH or credential.helper store), made workspace path configurable via CLAWNG_WORKSPACE env var instead of hardcoded path.
v1.0.1
Updated description: emphasizes portability (restore your agent on any machine via git clone) and evolution tracking (watch your assistant grow over time through commit history).
v1.0.0
Initial release — git-based long-term memory for OpenClaw agents. Tracks SOUL.md, MEMORY.md, and all core knowledge files with automatic GitHub push.
元数据
Slug clawng-term-memory
版本 1.2.0
许可证
累计安装 0
当前安装数 0
历史版本数 5
常见问题

Clawng Term Memory 是什么?

Make your OpenClaw agent portable and persistent. Version-controls SOUL.md, MEMORY.md, and all core knowledge files using git with automatic GitHub push — so... 它是一个面向 Claude Code / OpenClaw 的 AI Agent Skill 插件,目前累计下载 549 次。

如何安装 Clawng Term Memory?

在 OpenClaw 或 Claude Code 对话框中运行命令「/install clawng-term-memory」即可一键安装,无需额外配置。

Clawng Term Memory 是免费的吗?

是的,Clawng Term Memory 完全免费(开源免费),可自由下载、安装和使用。

Clawng Term Memory 支持哪些平台?

Clawng Term Memory 跨平台运行,可在任意部署了 OpenClaw / Claude Code 的环境中使用(cross-platform)。

谁开发了 Clawng Term Memory?

由 davidharket(@davidharket)开发并维护,当前版本 v1.2.0。

💬 留言讨论