# Scout Agent Deployment Checklist

## Pre-Deployment

### Configuration Review
- [ ] SOUL.md reviewed and approved
- [ ] HEARTBEAT.md schedule confirmed (every 2 hours)
- [ ] AGENTS.md workspace rules understood
- [ ] sources.json configured with preferred sources
- [ ] keywords.json includes RateRight-specific terms
- [ ] telegram.json has correct group ID (-1003505625266)

### Telegram Setup
- [ ] Create @ScoutRateRightBot (or similar name)
- [ ] Add bot to Rocky & Rivet group
- [ ] Grant message posting permissions
- [ ] Test bot can send to Research topic (ID: 5)
- [ ] Store bot token in Clawdbot secrets

### API Access (Optional but Recommended)
- [ ] Twitter/X API access for enhanced monitoring
- [ ] ArXiv API (free, no key required)
- [ ] Hacker News API (free, no key required)
- [ ] TechCrunch RSS feed access

## Deployment Steps

### 1. Activate Scout Agent
```bash
# Navigate to agent directory
cd /home/ccuser/rateright-growth/rivet/agents/scout

# Initialize Scout in Clawdbot
clawdbot agent create scout --type=monitoring --persona=SOUL.md

# Configure heartbeat schedule
clawdbot agent scout heartbeat --schedule="0 */2 * * *" --config=HEARTBEAT.md
```

### 2. Load Configuration
```bash
# Import source configuration
clawdbot agent scout config load sources.json

# Import keyword filters
clawdbot agent scout config load keywords.json

# Import Telegram settings
clawdbot agent scout config load telegram.json
```

### 3. Test Integration
```bash
# Test source connectivity
clawdbot agent scout test --sources=all --verbose

# Test Telegram messaging
clawdbot agent scout send-test --type=daily --topic=research

# Test urgent alert
clawdbot agent scout send-test --type=urgent --topic=research
```

### 4. Initialize Monitoring
```bash
# Run first scan manually
clawdbot agent scout scan --sources=all --output=console

# Review and approve filter performance
clawdbot agent scout analyze --date=today --show-scores

# Start automated monitoring
clawdbot agent scout start --heartbeat --logs=logs/scout.log
```

## Post-Deployment Verification

### First Day
- [ ] Confirm 2-hour heartbeat is running
- [ ] Verify daily briefing sent at 17:00 AEDT
- [ ] Check logs for any errors
- [ ] Test urgent alert with high-priority keyword

### First Week
- [ ] Review daily briefing quality
- [ ] Adjust keyword filters based on results
- [ ] Add any missing sources
- [ ] Calibrate relevance scoring

### First Month
- [ ] Analyze pattern detection accuracy
- [ ] Review weekly summary usefulness
- [ ] Update competitor list
- [ ] Gather team feedback on relevance

## Maintenance Schedule

### Daily
- Monitor logs for errors
- Verify alerts are working
- Check source availability

### Weekly
- Review and update keywords
- Archive old reports
- Check source reliability scores
- Update filtering criteria

### Monthly
- Add new sources
- Remove dead/irrelevant sources
- Update competitor tracking
- Review SOUL.md for accuracy

## Troubleshooting Commands

```bash
# Check agent status
clawdbot agent scout status

# View recent logs
clawdbot agent scout logs --tail=100

# Pause monitoring
clawdbot agent scout pause

# Resume monitoring
clawdbot agent scout resume

# Update configuration
clawdbot agent scout config reload

# Manual scan with filters
clawdbot agent scout scan --keywords=high --output=report.md
```

## Success Metrics

### Week 1 Targets
- [ ] 7 daily briefings delivered
- [ ] 0 false urgent alerts
- [ ] All sources responding
- [ ] Team feedback score >7/10

### Month 1 Targets
- [ ] Pattern detection working
- [ ] Weekly summaries valuable
- [ ] No downtime >30 minutes
- [ ] 90% of alerts relevant

## Emergency Contacts

- Technical Issues: Development team
- Source API Problems: Document in logs
- Telegram Issues: Check bot permissions
- Content Quality: Michael for review

---

*Deploy Scout v1.0 - AI News Intelligence for RateRight*