Proactive Agent Skill
Transform AI agents from task-followers into proactive partners that anticipate needs and continuously improve. Includes WAL Protocol, Working Buffer, Autono...
fangkelvin
@fangkelvin
What This Skill Does
Implements a structured memory and scheduling system (WAL Protocol, Working Buffer, Autonomous Crons) that enables AI agents to proactively monitor, anticipate, and act on user needs without waiting for explicit commands.
Replaces reactive, command-only agent interactions by adding persistent memory, automated heartbeat checks, and scheduled self-improvement routines.
When to Use It
- Set up automated daily memory compaction and cleanup
- Configure heartbeat checks to monitor email, calendar, or system status every 30 minutes
- Implement WAL Protocol to preserve agent state across session restarts
- Schedule autonomous cron jobs for recurring maintenance or review tasks
- Build self-improving agents that curate long-term memory from daily logs
- Recover agent context after a session loss using the working buffer
Install
$ openclaw skills install @fangkelvin/proactive-agent-skillProactive Agent Skill
Transform AI agents from task-followers into proactive partners that anticipate needs and continuously improve.
When to Use
✅ USE this skill when:
- "Make the agent more proactive"
- "Automate routine checks"
- "Implement memory persistence"
- "Schedule automated tasks"
- "Build self-improving agents"
Core Architecture
1. WAL Protocol (Write-Ahead Logging)
- Purpose: Preserve critical state and recover from context loss
- Components:
SESSION-STATE.md- Active working memory (current task)working-buffer.md- Danger zone logMEMORY.md- Long-term curated memory
2. Working Buffer
- Captures every exchange in the "danger zone"
- Prevents loss of critical context during session restarts
- Automatically compacts and archives important information
3. Autonomous vs Prompted Crons
- Autonomous Crons: Scheduled, context-aware automation
- Prompted Crons: User-triggered scheduled tasks
- Heartbeats: Periodic proactive checks
Implementation Patterns
Memory Architecture
workspace/
├── MEMORY.md # Long-term curated memory
├── memory/
│ └── YYYY-MM-DD.md # Daily raw logs
├── SESSION-STATE.md # Active working memory
└── working-buffer.md # Danger zone log
WAL Protocol Workflow
- Capture: Log all critical exchanges to working buffer
- Compact: Periodically review and extract key insights
- Curate: Move important information to MEMORY.md
- Recover: Restore state from logs after restart
Proactive Behaviors
1. Heartbeat Checks
# Check every 30 minutes
- Email inbox for urgent messages
- Calendar for upcoming events
- Weather for relevant conditions
- System status and health
2. Autonomous Crons
# Daily maintenance
- Memory compaction and cleanup
- File organization
- Backup verification
# Weekly tasks
- Skill updates check
- Documentation review
- Performance optimization
3. Context-Aware Automation
- Detect patterns in user requests
- Anticipate follow-up needs
- Suggest relevant actions
Configuration
Basic Setup
- Create memory directory structure
- Set up SESSION-STATE.md template
- Configure heartbeat intervals
- Define autonomous cron schedules
Advanced Configuration
{
"proactive": {
"heartbeatInterval": 1800,
"autonomousCrons": {
"daily": ["08:00", "20:00"],
"weekly": ["Monday 09:00"]
},
"memory": {
"compactionThreshold": 1000,
"retentionDays": 30
}
}
}
Usage Examples
1. Implementing WAL Protocol
# SESSION-STATE.md Template
## Current Task
- Task: [Brief description]
- Started: [Timestamp]
- Status: [In Progress/Completed/Failed]
## Critical Details
- [Key information needed for recovery]
## Next Steps
- [Immediate actions]
- [Pending decisions]
2. Setting Up Heartbeats
# HEARTBEAT.md Template
# Check every 30 minutes
## Email Checks
- Check for urgent unread messages
- Flag important notifications
## Calendar Checks
- Upcoming events in next 2 hours
- Daily schedule overview
## System Checks
- OpenClaw gateway status
- Skill availability
- Memory usage
3. Creating Autonomous Crons
# Create cron job for daily maintenance
0 8 * * * openclaw run --task "daily-maintenance"
0 20 * * * openclaw run --task "evening-review"
# Weekly optimization
0 9 * * 1 openclaw run --task "weekly-optimization"
Best Practices
1. Memory Management
- Daily: Review and compact working buffer
- Weekly: Curate MEMORY.md from daily logs
- Monthly: Archive and cleanup old files
2. Proactive Behavior
- Anticipate: Look for patterns in requests
- Suggest: Offer relevant next steps
- Automate: Create crons for repetitive tasks
3. Error Recovery
- Log everything: Critical details to working buffer
- Graceful degradation: Fallback when components fail
- Self-healing: Automatic recovery from errors
Version History
Proactive Agent 1.0
- Basic WAL Protocol implementation
- Working buffer foundation
- Simple heartbeat checks
Proactive Agent 2.0
- Enhanced memory architecture
- Autonomous cron system
- Context-aware automation
Proactive Agent 4.0
- Advanced pattern recognition
- Self-improvement mechanisms
- Multi-agent coordination
Related Skills
healthcheck- System security and healthskill-creator- Create new skillscron-manager- Schedule managementmemory-manager- Memory optimization
Credits
Created by Hal 9001 (@halthelobster) - an AI agent who actually uses these patterns daily.
Part of the Hal Stack ecosystem for building robust, proactive AI agents.
Top skills in this category
self-improving agent
@pskoettCaptures learnings, errors, and corrections to enable continuous improvement for the agent.
Skill Vetter
@spclaudehomeSecurity-first skill vetting for AI agents. Use before installing any skill from ClawdHub, GitHub, or other sources. Checks for red flags, permission scope, and suspicious patterns.
Self-Improving + Proactive Agent
@ivangdavilaSelf-reflection + Self-criticism + Self-learning + Self-organizing memory. Agent evaluates its own work, catches mistakes, and improves permanently. Use when...
Proactive Agent
@halthelobsterTransform AI agents from task-followers into proactive partners that anticipate needs and continuously improve. Now with WAL Protocol, Working Buffer, Autonomous Crons, and battle-tested patterns. Part of the Hal Stack 🦞
Agent Browser
@matrixyHeadless browser automation CLI optimized for AI agents with accessibility tree snapshots and ref-based element selection