Proactive Memory Agent — AI 智能体终极记忆优化系统

Name: Proactive Memory Agent — AI 智能体终极记忆优化系统
Author: silaszhu

silaszhu

Proactive Memory Agent — AI 智能体终极记忆优化系统

v1.0.1

该技能为 AI 智能体提供终极记忆优化，结合 WAL 预写协议、工作缓冲区、三级记忆架构（HOT/WARM/COLD）和上下文预算（10/40/20/20），实现 60-70% 的上下文占用降低，零信息丢失，适用于长会话、复杂任务、高频交互和成本控制场景。

0· 95·0 当前·0 累计

by @silaszhu·MIT-0

开发工具 API工具自动化 AI模型访问

下载技能包

License

MIT-0

最后更新

2026/3/24

安全扫描

VirusTotal

Pending

查看报告

OpenClaw

安全

high confidence

技能文件和运行指令内部一致，用于本地记忆管理，不请求无关凭证或进行网络数据泄漏，但会读取和快照 OpenClaw 工作空间下的任意文件，请在使用前审查工作空间内容。

评估建议

技能如描述，进行本地记忆分层和检查点管理。使用前：（1）自行检查脚本；（2）确保工作空间无敏感文件；（3）了解可选工具；（4）考虑在受限环境运行或备份；（5）验证无远程端点添加。...

详细分析 ▾

✓ 用途与能力

Name/description (memory optimization, WAL, tiering, context budgeting) match the included scripts and SKILL.md. The scripts only create and manage files under the OpenClaw workspace (SESSION-STATE.md, .learnings/, memory/ etc.), and optionally call local tools like jq or the openclaw CLI. Nothing required by the skill (no env vars, no credentials, no external binaries) is inconsistent with its stated purpose.

ℹ 指令范围

Runtime instructions direct the agent/operator to run init.sh, detect.sh, search.sh, tiering.sh, and checkpoint.sh which read, write, copy and archive workspace files. This is expected for a memory manager, but the scripts aggregate and snapshot content (e.g., copying SESSION-STATE.md into snapshots), so they will capture any text present in your workspace files. There are no network endpoints or remote uploads in the scripts, but the scope of files touched is broad within the workspace.

✓ 安装机制

No install spec or external downloads are used. The skill is instruction-plus-scripts only; all logic is present in the included shell scripts. This is lower risk than remote installers. The scripts do rely optionally on local utilities (jq, openclaw CLI) if present.

ℹ 凭证需求

The skill declares no required environment variables or credentials. It does respect an OPENCLAW_WORKSPACE variable if set and otherwise uses $HOME/.openclaw/workspace. While that is proportional to its purpose, the scripts will read and archive files under that workspace — which could include tokens or secrets if you store them there. No unrelated external credentials are requested.

✓ 持久化与权限

The skill does not request always:true, does not modify other skills or system-wide configs, and only creates files under the workspace path. Its persistence is limited to workspace files it creates/updates (.proactive-memory-state.json, SESSION-STATE.md, memory/ and .learnings/).

安全有层次，运行前请审查代码。

License

MIT-0

可自由使用、修改和再分发，无需署名。

查看条款 ↗

运行时依赖

无特殊依赖

版本

latestv1.0.12026/3/24

v1.0.1：添加全面中文文档支持，SKILL.md 双语支持，包括完整中文翻译和示例。

● Pending

安装命令点击复制

官方npx clawhub@latest install proactive-memory-agent

镜像加速npx clawhub@latest install proactive-memory-agent --registry https://cn.clawhub-mirror.com

技能文档

（由于原始内容过长，以下仅提供部分翻译，保留原始 Markdown 格式和代码块）

The ultimate memory optimization system for OpenClaw agents.

OpenClaw 智能体终极记忆优化系统。

Combines the best of three worlds:

🦞 Proactive Agent — WAL protocol, Working Buffer, learning system
📊 Context Budgeting — Strict 10/40/20/20 partitioning
🗂️ Memory Tiering — Automatic HOT/WARM/COLD archiving

Result: 60-70% context reduction + complete learning retention + zero maintenance

效果：降低60-70%上下文占用 + 完整保留学习记录 + 零维护成本

快速开始 / Quick Start

中文简介

Proactive Memory Agent 是一个专为 OpenClaw 设计的记忆优化系统，解决长会话中的上下文爆炸问题。

核心特性：

🔥 WAL 预写协议 — 关键决策先写后回，确保信息不丢失
⚠️ 工作缓冲区 — 危险区(>60%)自动记录所有对话
📊 上下文预算 — 严格 10/40/20/20 分区控制
🗂️ 三级记忆架构 — HOT→WARM→COLD 自动归档
📝 学习记录系统 — 自动记录错误、纠正和最佳实践

适用场景：

长会话 (>50条消息)
复杂任务 (多步骤、多文件)
高频交互 (需要记住用户偏好)
成本控制 (降低token消耗)

预期效果：

Context 占用降低 60-70%
Token 消耗降低 50-60%
信息丢失风险 几乎为零

Memory Architecture / 记忆架构

workspace/
├── SESSION-STATE.md          # 🔥 HOT: Current task (WAL target)
├── .learnings/               # 🌡️ WARM: Learning records
│   ├── ERRORS.md            — 错误记录
│   ├── LEARNINGS.md         — 学习/纠正记录
│   └── FEATURE_REQUESTS.md  — 功能请求记录
├── MEMORY.md                 # ❄️ COLD: Long-term archive
├── memory/
│   ├── hot/HOT_MEMORY.md     # 快速恢复检查点
│   ├── warm/                 # 稳定偏好/配置
│   ├── cold/                 # 长期归档
│   └── working-buffer.md     # ⚠️ 危险区日志
└── AGENTS.md/SOUL.md/TOOLS.md — 只读参考

The Four Protocols / 四大协议

1️⃣ WAL Protocol (Write-Ahead Logging) / 预写协议

Law: Context is a BUFFER, not storage. Write first, respond second. 法则： 上下文是缓冲，不是存储。先写再回。

SCAN every message for: 扫描每条消息中的：

✏️ Corrections — "It's X, not Y" / "Actually..." / 纠正
📍 Proper nouns — Names, places, products / 专有名词
🎨 Preferences — Colors, styles, approaches / 偏好
📋 Decisions — "Let's do X" / "Go with Y" / 决策
🔢 Specific values — Numbers, dates, IDs, URLs / 具体数值

The Protocol:

STOP — Do not start composing your response / 停止，不要开始回复
WRITE — Update SESSION-STATE.md with the detail / 写入 SESSION-STATE.md
THEN — Respond to your human / 然后回复用户

2️⃣ Working Buffer Protocol / 工作缓冲区协议

Purpose: Capture EVERY exchange in the danger zone (>60% context). 目的： 在危险区(>60%上下文)记录每一条消息。

中文说明： 当上下文使用率超过 60% 时，系统自动启用工作缓冲区，记录每一条对话。即使发生压缩，也能从缓冲区恢复。

How:

At 60% context: Clear old buffer, start fresh
Every message after 60%: Append to memory/working-buffer.md
After compaction: Read buffer FIRST, extract important context
Leave buffer as-is until next 60% threshold

3️⃣ Context Budgeting (10/40/20/20) / 上下文预算分区

Strict partitioning / 严格分区：

Zone	%	Content	File	中文说明
Objective	10%	Core task, active constraints	SESSION-STATE.md	核心任务
Short-term	40%	Recent 5-10 turns raw dialogue	Working Buffer	短期对话
Decision Log	20%	Summarized outcomes	.learnings/.md	决策日志
Background	20%	High-relevance snippets	MEMORY.md	背景知识

Rule: When any zone exceeds quota, oldest content moves to lower tier. 规则： 任何分区超出配额时，最旧内容移至下一层级。

4️⃣ Memory Tiering (HOT → WARM → COLD) / 记忆分层归档

Automatic lifecycle / 自动生命周期：

🔥 HOT (Current session) 当前会话
   ↓ (Task complete) 任务完成
🌡️ WARM (Recurring use) 重复使用
   ↓ (30 days old / promoted) 30天旧/已提升
❄️ COLD (Archive) 归档

Tier	Location	Update Frequency	Retention	中文
HOT	SESSION-STATE.md	Every message	Current task only	当前任务
WARM	.learnings/	When learning occurs	Until promoted	学习记录
COLD	MEMORY.md	Weekly archival	Permanent	长期存档

使用脚本 / Scripts

快速设置 / Quick Setup

~/.openclaw/workspace/skills/proactive-memory-agent/scripts/init.sh

日常使用 / Daily Operations

# 检测当前状态 / Check current status ~/.openclaw/workspace/skills/proactive-memory-agent/scripts/detect.sh # 搜索记忆 / Search memories ~/.openclaw/workspace/skills/proactive-memory-agent/scripts/search.sh "关键词" # 执行记忆分层归档 / Run memory tiering ~/.openclaw/workspace/skills/proactive-memory-agent/scripts/tiering.sh

# 创建预压缩检查点 / Create pre-compaction checkpoint ~/.openclaw/workspace/skills/proactive-memory-agent/scripts/checkpoint.sh

学习记录系统 / Learning System (.learnings/)

何时记录 / When to Record

场景 / Situation	记录位置 / Record To	类别 / Category
命令失败 / Command fails	ERRORS.md	error
用户纠正 / User corrects	LEARNINGS.md	correction
缺失功能 / Missing feature	FEATURE_REQUESTS.md	feature_request
知识过时 / Knowledge outdated	LEARNINGS.md	knowledge_gap
更好方法 / Better approach	LEARNINGS.md	best_practice

记录格式 / Format

## [LRN-YYYYMMDD-XXX] category
Logged: ISO-8601 timestamp
Priority: low | medium | high | critical
Status: pending | in_progress | resolved | promoted
Summary / 摘要
一句话描述学到了什么
Details / 详情
What happened, what was wrong, correct approach
发生了什么，哪里错了，正确做法
Suggested Action / 建议操作
Specific fix or improvement
具体的修复或改进建议

提升路径 / Promotion Path

Learning in .learnings/ 在学习记录中
    ↓
IF recurring (3+ times) OR broadly applicable 如果重复(3+次)或广泛适用
    ↓
Promote to / 提升到:
  ├─ Behavioral patterns → SOUL.md 行为模式
  ├─ Workflow improvements → AGENTS.md 工作流改进
  └─ Tool gotchas → TOOLS.md 工具技巧

心跳集成 / Heartbeat Integration

Add to HEARTBEAT.md: 添加到心跳文件：

## Memory Management / 记忆管理 (每30分钟)
[ ] Run detect.sh — 检查context使用率
[ ] If >80%: Create checkpoint with checkpoint.sh / 创建检查点
[ ] If >70%: Run tiering.sh to archive old content / 归档旧内容
[ ] Review .learnings/ for pending items / 检查待处理项

核心原则 / Key Principles

Write before respond — WAL is non-negotiable / 先写再回，WAL不可妥协
Buffer the danger zone — >60% = automatic logging / 缓冲危险区，>60%自动记录
Strict budgeting — 10/40/20/20, no exceptions / 严格预算，10/40/20/20无例外
Auto-tiering — HOT→WARM→COLD without manual effort / 自动分层，无需手动
Promote aggressively — Recurring patterns become permanent / 积极提升，重复模式永久化
Never lose context — Working Buffer + Compaction Recovery / 永不丢失，缓冲+恢复

预期效果 / Expected Results

Metric / 指标	Before / 之前	After / 之后	Improvement / 改进
Context usage / 上下文占用	100%	30-40%	60-70% reduction
Token consumption / Token消耗	Baseline / 基准	-50-60%	Major savings
Maintenance / 维护成本	High / 高	Minimal / 极低	Auto-tiering
Information loss / 信息丢失	Risk / 有风险	Zero / 零	Buffer protection

Changelog / 更新日志

v1.0.1 (2026-03-24)

Added Chinese documentation / 添加中文文档
Improved README with bilingual support / 改进双语支持
Fixed minor formatting issues / 修复格式问题

v1.0.0 (2026-03-24)

Initial release / 初始版本
Integrated: proactive-agent + context-budgeting + memory-tiering / 整合三大技能
5 management scripts included / 包含5个管理脚本
Full documentation in SKILL.md / 完整文档

License / 许可证

MIT License — Free to use, modify, distribute. MIT 许可证 — 自由使用、修改、分发。

Part of the Hal Stack 🦞 + Self-Improvement Integration Integrated by @silaszhu*

数据来源：ClawHub ↗ · 中文优化：龙虾技能库

OpenClaw 技能定制 / 插件定制 / 私有工作流定制

免费技能或插件可能存在安全风险，如需更匹配、更安全的方案，建议联系付费定制

了解定制服务

License

运行时依赖

版本

安装命令 点击复制

技能文档

快速开始 / Quick Start

中文简介

Memory Architecture / 记忆架构

The Four Protocols / 四大协议

1️⃣ WAL Protocol (Write-Ahead Logging) / 预写协议

2️⃣ Working Buffer Protocol / 工作缓冲区协议

3️⃣ Context Budgeting (10/40/20/20) / 上下文预算分区

4️⃣ Memory Tiering (HOT → WARM → COLD) / 记忆分层归档

使用脚本 / Scripts

快速设置 / Quick Setup

日常使用 / Daily Operations

学习记录系统 / Learning System (.learnings/)

何时记录 / When to Record

记录格式 / Format

Summary / 摘要

Details / 详情

Suggested Action / 建议操作

提升路径 / Promotion Path

心跳集成 / Heartbeat Integration

核心原则 / Key Principles

预期效果 / Expected Results

Changelog / 更新日志

v1.0.1 (2026-03-24)

v1.0.0 (2026-03-24)

License / 许可证

安装命令点击复制