🧪 Skills

Session Guardian

对话永不丢失，任务永不混淆。解决模型掉线、Gateway重启、跨渠道混淆、任务难追踪等痛点。五层防护：增量备份（5分钟）+ 快照（1小时）+ 智能总结

v3.0.0
⭐ —
❤️ 1
⬇️ 254
👁 1
Save📁 Collect
Share

Description

name: session-guardian description: "Never lose a conversation again. Auto-backup every 5 minutes, smart summaries, and instant recovery when models crash or Gateway restarts. Works for solo agents and multi-agent teams. Use when you need bulletproof session protection, task tracking across agents, or collaboration history. Prevents token overflow with AI-powered compaction (30-50% savings)." metadata:

  • multi-agent
  • data-protection
  • automation metadata: openclaw: emoji: "🛡️" minVersion: "0.9.0"

Session Guardian 🛡️

Your Conversation Guardian - Enterprise-grade session backup + project management solution

Problem Statement

  • 🔴 Model disconnections - Conversations lost, work wasted
  • 🔴 Gateway restarts - Forgot what you were doing, task state lost
  • 🔴 Cross-channel confusion - Private info sent to group chat, or vice versa
  • 🔴 Complex task tracking - Tasks span multiple sessions, hard to track state
  • 🔴 Multi-agent chaos - Multiple agents working simultaneously, unclear who's doing what
  • 🔴 Large session files - Causing timeouts, slow responses, high token consumption

Quick Start

# Install
clawhub install session-guardian

# One-click deployment (auto-configure all cron jobs)
cd ~/.openclaw/workspace/skills/session-guardian
bash scripts/install.sh

# Verify
crontab -l | grep session-guardian
openclaw cron list

Use Cases

Scenario 1: Enterprise Multi-Agent Collaboration

Typical User: Enterprise teams with multiple agents working together

Example:

  • Main agent → Dev team lead → UI designer + Full-stack developer
  • Main agent → Finance team lead → Strategy research + Risk control + Execution
  • Main agent → Operations team lead → Content creation + Data analysis

Benefits:

  • ✅ Collaboration tracking: Visualize task flow
  • ✅ Collaboration health: Monitor team collaboration quality
  • ✅ Knowledge base: Accumulate team best practices

Scenario 2: Personal Assistant Team

Typical User: Individual users with multiple specialized assistants

Example:

  • Research assistant: Find resources, organize literature
  • Coding assistant: Write code, debug issues
  • Writing assistant: Write articles, polish copy
  • Translation assistant: Multi-language translation

Benefits:

  • ✅ Smart backup: Protect each assistant's memory
  • ✅ Knowledge extraction: Each assistant accumulates expertise
  • ✅ Collaboration tracking: Know which assistant has the task

Scenario 3: Single-Agent Deep Usage

Typical User: Users with one main agent

Example:

  • Personal AI assistant: Handle all daily tasks
  • Professional consultant: Deep conversations, long-term memory

Benefits:

  • ✅ Smart backup: Prevent conversation loss
  • ✅ Knowledge extraction: Accumulate personal best practices
  • ✅ Task management: Track complex task states

Scenario 4: Enterprise Multi-Department Collaboration

Typical User: Enterprises with agents in different departments

Example:

  • Sales agent: Customer management, opportunity tracking
  • Support agent: Issue handling, ticket management
  • Tech agent: Technical support, troubleshooting
  • Finance agent: Report generation, data analysis

Benefits:

  • ✅ Session isolation: Prevent cross-department information leakage
  • ✅ Collaboration health: Monitor cross-department collaboration

Core Features

1. Five-Layer Protection System

Layer 1: Incremental Backup (Every 5 minutes)

  • Auto-backup new conversations
  • Minimal storage, fast recovery
  • Supports all agent sessions

Layer 2: Snapshot Backup (Every 1 hour)

  • Complete session snapshots
  • Version control, rollback support
  • Automatic cleanup of old versions

Layer 3: Smart Summary (Daily)

  • Extract key information from conversations
  • Auto-update MEMORY.md
  • Generate collaboration summaries

Layer 4: Health Check (Every 6 hours)

  • Monitor session file size
  • Detect abnormal token consumption
  • Alert on collaboration issues

Layer 5: Project Management

  • Task state tracking
  • Milestone management
  • Progress reports

2. Multi-Agent Collaboration Features (v2.0)

Collaboration Tracking

  • Visualize task flow: King → Team Lead → Members
  • Record collaboration links
  • Generate collaboration reports

Smart Backup Strategy

  • High-frequency agents: More frequent backups
  • Low-frequency agents: Reduce backup frequency
  • Save storage and token consumption

Knowledge Base Extraction

  • Extract best practices from conversations
  • Auto-update AGENTS.md
  • Generate FAQ

Collaboration Health Scoring

  • Communication efficiency
  • Task completion rate
  • Response time
  • Collaboration quality

3. Integration with Self-Improving-Agent

Complementary Design:

  • Session Guardian: Macro perspective, overall progress
  • Self-Improving-Agent: Micro perspective, specific issues

Workflow:

  1. Self-Improving-Agent records errors and learnings in real-time (lightweight)
  2. Session Guardian reads SI records during daily summary
  3. Session Guardian extracts knowledge to MEMORY.md
  4. Avoid duplicate analysis

Token Savings:

  • Before: ~30k tokens/day
  • After: ~10.5k tokens/day
  • Save 65% tokens

Installation

Method 1: ClawHub (Recommended)

clawhub install session-guardian

Method 2: Manual Installation

# Clone repository
git clone https://github.com/1052326311/session-guardian.git ~/.openclaw/workspace/skills/session-guardian

# Run installation script
cd ~/.openclaw/workspace/skills/session-guardian
bash scripts/install.sh

Method 3: npm Installation

npm install -g @openclaw/session-guardian

Configuration

Basic Configuration

Edit ~/.openclaw/workspace/skills/session-guardian/config.json:

{
  "backup": {
    "incremental": {
      "enabled": true,
      "interval": 5,
      "retention": 7
    },
    "snapshot": {
      "enabled": true,
      "interval": 60,
      "retention": 30
    }
  },
  "summary": {
    "enabled": true,
    "schedule": "0 0 * * *"
  },
  "healthCheck": {
    "enabled": true,
    "interval": 6
  },
  "collaboration": {
    "tracking": true,
    "healthScore": true
  },
  "integration": {
    "selfImprovingAgent": {
      "enabled": true,
      "readLearnings": true
    }
  }
}

Advanced Configuration

Custom Backup Path

{
  "backup": {
    "path": "/custom/backup/path"
  }
}

Agent-Specific Settings

{
  "agents": {
    "main": {
      "backup": {
        "interval": 3
      }
    },
    "dev-lead": {
      "backup": {
        "interval": 10
      }
    }
  }
}

Usage

Manual Backup

# Backup all sessions
openclaw skill run session-guardian backup

# Backup specific agent
openclaw skill run session-guardian backup --agent main

Manual Summary

# Generate daily summary
openclaw skill run session-guardian summary

# Generate weekly summary
openclaw skill run session-guardian summary --weekly

Health Check

# Run health check
openclaw skill run session-guardian health-check

# View health report
cat ~/.openclaw/workspace/Assets/SessionBackups/health-reports/latest.md

Restore Session

# List available backups
openclaw skill run session-guardian list-backups

# Restore from backup
openclaw skill run session-guardian restore --backup 2026-03-06-10-00

File Structure

~/.openclaw/workspace/
├── skills/
│   └── session-guardian/
│       ├── SKILL.md                    # This file
│       ├── config.json                 # Configuration
│       ├── scripts/
│       │   ├── install.sh              # Installation script
│       │   ├── incremental-backup.sh   # Incremental backup
│       │   ├── snapshot-backup.sh      # Snapshot backup
│       │   ├── daily-summary.sh        # Daily summary
│       │   ├── health-check.sh         # Health check
│       │   ├── collaboration-tracker.sh # Collaboration tracking
│       │   └── knowledge-extractor.sh  # Knowledge extraction
│       └── templates/
│           ├── summary.md              # Summary template
│           └── health-report.md        # Health report template
└── Assets/
    └── SessionBackups/
        ├── incremental/                # Incremental backups
        ├── snapshots/                  # Snapshot backups
        ├── summaries/                  # Daily summaries
        ├── health-reports/             # Health reports
        └── collaboration/              # Collaboration data

Cron Jobs

After installation, the following cron jobs are automatically configured:

# Incremental backup (every 5 minutes)
*/5 * * * * cd ~/.openclaw/workspace/skills/session-guardian && bash scripts/incremental-backup.sh

# Snapshot backup (every hour)
0 * * * * cd ~/.openclaw/workspace/skills/session-guardian && bash scripts/snapshot-backup.sh

# Daily summary (midnight)
0 0 * * * cd ~/.openclaw/workspace/skills/session-guardian && bash scripts/daily-summary.sh

# Health check (every 6 hours)
0 */6 * * * cd ~/.openclaw/workspace/skills/session-guardian && bash scripts/health-check.sh

# Collaboration tracking (every 30 minutes)
*/30 * * * * cd ~/.openclaw/workspace/skills/session-guardian && bash scripts/collaboration-tracker.sh

Best Practices

1. Regular Monitoring

  • Check health reports weekly
  • Review collaboration health scores
  • Monitor token consumption

2. Backup Management

  • Keep 7 days of incremental backups
  • Keep 30 days of snapshots
  • Archive important conversations

3. Knowledge Extraction

  • Review daily summaries
  • Update MEMORY.md regularly
  • Share best practices with team

4. Integration with Other Skills

  • Use with self-improving-agent for error learning
  • Use with project management tools
  • Integrate with notification systems

Troubleshooting

Backup Not Running

# Check cron jobs
crontab -l | grep session-guardian

# Check logs
tail -f ~/.openclaw/workspace/skills/session-guardian/logs/backup.log

Large Session Files

# Run health check
openclaw skill run session-guardian health-check

# View recommendations
cat ~/.openclaw/workspace/Assets/SessionBackups/health-reports/latest.md

Missing Summaries

# Manually generate summary
openclaw skill run session-guardian summary

# Check summary logs
tail -f ~/.openclaw/workspace/skills/session-guardian/logs/summary.log

Performance

Token Consumption

  • Incremental backup: ~100 tokens/run
  • Snapshot backup: ~500 tokens/run
  • Daily summary: ~5k tokens/run (with SI integration: ~8k)
  • Health check: ~200 tokens/run
  • Collaboration tracking: ~300 tokens/run

Total: ~10-15k tokens/day (with SI integration)

Storage

  • Incremental backup: ~10KB/session/day
  • Snapshot backup: ~100KB/session/day
  • Summaries: ~5KB/day
  • Health reports: ~2KB/report

Total: ~1-2MB/agent/month

Roadmap

v2.1 (Planned)

  • Web dashboard for visualization
  • Real-time collaboration monitoring
  • Advanced analytics
  • Custom alert rules

v2.2 (Planned)

  • Cloud backup support
  • Team collaboration features
  • API for third-party integration
  • Mobile app

Contributing

Contributions are welcome! Please:

  1. Fork the repository
  2. Create a feature branch
  3. Submit a pull request

License

MIT License - see LICENSE file for details

Support

Changelog

v2.0.0 (2026-03-05)

  • ✨ Added collaboration tracking
  • ✨ Added smart backup strategy
  • ✨ Added knowledge extraction
  • ✨ Added collaboration health scoring
  • ✨ Integration with self-improving-agent
  • 🐛 Fixed session isolation issues
  • 📝 Improved documentation

v1.0.0 (2026-03-01)

  • 🎉 Initial release
  • ✨ Five-layer protection system
  • ✨ Multi-agent support
  • ✨ Project management features

中文版 / Chinese Version

核心功能

1. 五层防护体系

第1层：增量备份（每5分钟）

  • 自动备份新对话
  • 存储占用小，恢复快
  • 支持所有agent的session

第2层：快照备份（每1小时）

  • 完整session快照
  • 版本控制，支持回滚
  • 自动清理旧版本

第3层：智能总结（每天）

  • 从对话中提取关键信息
  • 自动更新MEMORY.md
  • 生成协作总结

第4层：健康检查（每6小时）

  • 监控session文件大小
  • 检测异常token消耗
  • 协作问题预警

第5层：项目管理

  • 任务状态追踪
  • 里程碑管理
  • 进度报告

2. 多智能体协作功能（v2.0）

协作链路追踪

  • 可视化任务流转：King → 团长 → 成员
  • 记录协作链路
  • 生成协作报告

智能备份策略

  • 高频agent：更频繁备份
  • 低频agent：降低备份频率
  • 节省存储和token消耗

知识库沉淀

  • 从对话中提取最佳实践
  • 自动更新AGENTS.md
  • 生成FAQ

协作健康度评分

  • 沟通效率
  • 任务完成率
  • 响应时间
  • 协作质量

3. 与Self-Improving-Agent集成

互补设计：

  • Session Guardian：宏观视角，看整体进展
  • Self-Improving-Agent：微观视角，看具体问题

工作流：

  1. Self-Improving-Agent实时记录错误和学习（轻量级）
  2. Session Guardian每日总结时读取SI记录
  3. Session Guardian提取知识到MEMORY.md
  4. 避免重复分析

Token节省：

  • 优化前：~30k tokens/天
  • 优化后：~10.5k tokens/天
  • 节省65% tokens

安装

方法1：ClawHub（推荐）

clawhub install session-guardian

方法2：手动安装

# 克隆仓库
git clone https://github.com/1052326311/session-guardian.git ~/.openclaw/workspace/skills/session-guardian

# 运行安装脚本
cd ~/.openclaw/workspace/skills/session-guardian
bash scripts/install.sh

配置

编辑 ~/.openclaw/workspace/skills/session-guardian/config.json:

{
  "backup": {
    "incremental": {
      "enabled": true,
      "interval": 5,
      "retention": 7
    },
    "snapshot": {
      "enabled": true,
      "interval": 60,
      "retention": 30
    }
  },
  "summary": {
    "enabled": true,
    "schedule": "0 0 * * *"
  },
  "healthCheck": {
    "enabled": true,
    "interval": 6
  },
  "collaboration": {
    "tracking": true,
    "healthScore": true
  },
  "integration": {
    "selfImprovingAgent": {
      "enabled": true,
      "readLearnings": true
    }
  }
}

使用

手动备份

# 备份所有session
openclaw skill run session-guardian backup

# 备份特定agent
openclaw skill run session-guardian backup --agent main

手动总结

# 生成每日总结
openclaw skill run session-guardian summary

# 生成每周总结
openclaw skill run session-guardian summary --weekly

健康检查

# 运行健康检查
openclaw skill run session-guardian health-check

# 查看健康报告
cat ~/.openclaw/workspace/Assets/SessionBackups/health-reports/latest.md

性能

Token消耗

  • 增量备份：~100 tokens/次
  • 快照备份：~500 tokens/次
  • 每日总结：~5k tokens/次（集成SI后：~8k）
  • 健康检查：~200 tokens/次
  • 协作追踪：~300 tokens/次

总计：~10-15k tokens/天（集成SI后）

存储

  • 增量备份：~10KB/session/天
  • 快照备份：~100KB/session/天
  • 总结：~5KB/天
  • 健康报告：~2KB/报告

总计：~1-2MB/agent/月

支持

更新日志

v2.0.0 (2026-03-05)

  • ✨ 新增协作链路追踪
  • ✨ 新增智能备份策略
  • ✨ 新增知识库沉淀
  • ✨ 新增协作健康度评分
  • ✨ 集成self-improving-agent
  • 🐛 修复session隔离问题
  • 📝 改进文档

v1.0.0 (2026-03-01)

  • 🎉 首次发布
  • ✨ 五层防护体系
  • ✨ 多智能体支持
  • ✨ 项目管理功能

Reviews (0)

Sign in to write a review.

No reviews yet. Be the first to review!

Comments (0)

Sign in to join the discussion.

No comments yet. Be the first to share your thoughts!

Compatible Platforms

Links

📂 Source Code

Pricing

Free

Related Configs

🧪 Skill

self-improving-agent

Free

Captures learnings, errors, and corrections to enable continuous improvement. Use when: (1) A command or operation fails unexpectedly, (2) User corrects Clau...

❤️ 2.0k ⬇️ 218k
🧪 Skill

Self Improving Agent

Free

Captures learnings, errors, and corrections to enable continuous improvement. And also 50+ models for image generation, video generation, text-to-speech, spe...

❤️ 2.0k ⬇️ 206k
🧪 Skill

Find Skills

Free

Search, discover, and install skills from the open agent skills ecosystem to extend agent capabilities for specific tasks or domains.

❤️ 814 ⬇️ 199k
🧪 Skill

Summarize

Free

--- name: summarize description: Summarize URLs or files with the summarize CLI (web, PDFs, images, audio, YouTube). homepage: https://summarize.sh metadata: {"clawdbot":{"emoji":"🧾","requires":{"b

❤️ 609 ⬇️ 160k