Daily Workflows¶

Common daily workflows for Hawkeye MCP users.

Morning Review Workflow¶

Check overnight incidents and prepare for the day.

Step 1: Check Overnight Incidents¶

Show me uninvestigated alerts from the last 12 hours

Step 2: Prioritize¶

Show me only P1 alerts from the last 12 hours

Link for #1
Link for #2

Share these links with your team or add to incident tickets.

Step 4: Investigate Critical¶

Investigate all P1 alerts

Step 5: Review Results¶

For each investigation, show me:
- Root cause
- Corrective actions
- Business impact

Step 6: Execute Fixes¶

Copy bash scripts and code snippets from RCAs and execute corrective actions.

Time: 15-20 minutes for 5-10 incidents

Incident Response Workflow¶

Active incident requiring immediate investigation.

Step 1: Find Latest Alert¶

Show me alerts from the last 30 minutes

Step 2: Immediate Investigation¶

Investigate the most recent alert and wait for results

Step 3: Quick RCA¶

Show me the root cause and corrective actions

Step 4: Execute Fix¶

Execute the first corrective action

Step 5: Verify¶

Check if the incident is resolved

Time: 2-5 minutes from alert to fix

Weekly Review Workflow¶

Analyze patterns and optimize instructions.

Step 1: Get Analytics¶

Show me incident statistics for the last 7 days

Step 2: Review Quality¶

Show me investigations with quality score < 80

Step 3: Identify Patterns¶

What are the most common incident types this week?

Step 4: Refine Instructions¶

For recurring incident types with low scores:

1. Review past RCAs
2. Draft new instruction
3. Test on past session
4. Deploy if improved

Time: 30-45 minutes weekly

Post-Deployment Workflow¶

Monitor new deployment for issues.

Step 1: Pre-Deployment Baseline¶

Show me current incident rate and types

Step 2: Deploy Application¶

Execute deployment via CI/CD.

Step 3: Monitor Post-Deployment (15 min)¶

Show me any new alerts in the last 15 minutes

Step 4: Quick Investigation¶

For any deployment-related alerts:
Investigate immediately
Check if deployment is the cause
Rollback if necessary

Step 5: Post-Deployment Report (1 hour)¶

Show me all incidents in the hour after deployment
Compare with baseline

Time: Continuous monitoring during deployment window

Monthly Optimization Workflow¶

Optimize Hawkeye configuration for better results.

Step 1: Monthly Analytics¶

Show me incident report for the last 30 days

Step 2: Review Instructions¶

List all active instructions
Identify unused or ineffective ones

Step 3: A/B Test New Approaches¶

1. Draft alternative instruction
2. Test on multiple past sessions
3. Compare results
4. Deploy winner, disable loser

Step 4: Review Connections¶

List all connections
Check sync status
Add new data sources if needed

Step 5: Update Context¶

Update SYSTEM instructions with:
- Architecture changes
- New services
- Updated SLAs

Time: 1-2 hours monthly

On-Call Workflow¶

Optimized workflow for on-call engineers.

1. Alert Notification¶

Receive alert from PagerDuty/OpsGenie.

2. Quick Check¶

Show me alerts for [service-name] in the last hour

3. Investigate¶

Investigate the alert and wait

4: Review RCA¶

Show me:
- Root cause
- Business impact
- Corrective actions

5: Take Action¶

If auto-remediated: Verify and close

If manual action needed: Execute provided bash scripts

If escalation needed: Page senior engineer with RCA summary

6: Document¶

Get link for investigation: Link for sessionID abc-123
Copy RCA to incident ticket
Add investigation link to ticket
Include time saved metric
Close incident

Time: 3-8 minutes per incident

Preventive Maintenance Workflow¶

Proactive issue prevention based on patterns.

Step 1: Identify Recurring Issues¶

Show me incidents grouped by type for last 30 days

Step 2: Extract Preventive Measures¶

For top 3 incident types:
Show me all preventive measures recommended

Step 3: Create Maintenance Tasks¶

Convert preventive measures to:
- Jira tickets
- Infrastructure improvements
- Process changes

Step 4: Track Implementation¶

Monitor if recurring issues decrease after fixes

Time: 1 hour monthly, saves hours in reduced incidents

New Team Member Onboarding¶

Training workflow for new engineers.

Day 1: Introduction¶

1. Show me all Hawkeye projects
2. Show me project details
3. Show me connections
4. Show me active instructions

Day 2: Read-Only Investigation¶

1. Show me past investigations
2. Review RCA examples
3. Ask follow-up questions
4. Learn RCA structure

Day 3: Practice Investigation¶

1. Find uninvestigated alert
2. Investigate with supervision
3. Review RCA quality
4. Execute safe corrective actions

Week 2: Independent Use¶

1. Investigate alerts independently
2. Create first instruction (with review)
3. Run weekly review

Time: 2 weeks to full proficiency

Performance Tracking Workflow¶

Track and report Hawkeye ROI.

Weekly Metrics¶

Show me:
- Total investigations this week
- Average MTTR
- Total time saved
- Quality score trend

Monthly Report¶

Generate monthly incident report:
- Investigations: Count, types, quality
- Performance: MTTR, time saved
- Noise reduction: Filtered vs investigated
- Trends: Week over week changes

Quarterly Review¶

Compare quarters:
- MTTR improvement
- Total time saved
- Cost avoidance (downtime prevented)
- Team efficiency gains

Time: 30 minutes weekly, 1 hour monthly, 2 hours quarterly