System watcher should also provide a leaderboard of consuming processes

[Gennadiyev]

Currently, the notification watches for CPU and RAM. When a server is down, admins have no idea why.

The system watcher is meant to provide real-time information for system admins to diagnose and find issues. A quick snapshot of what's running and what causes the extra strain can be really helpful.

[Gennadiyev]

@claude Can you try to solve this?

[claude]

Claude finished @Gennadiyev's task in 1m 55s —— View job • claude/issue-10-20260206-0417 • Create PR ➔

---

Added process leaderboard to system watcher ✅

Tasks

• Read repository guidelines (CLAUDE.md)
• Understand current system watcher implementation
• Implement process monitoring functionality
• Update notification formatting to include process leaderboard
• Commit and push changes
• Provide PR link

Changes Made

Added process leaderboard functionality to watchers/system_watcher.py:

New Methods:

• get_top_processes_by_cpu() - Fetches top 5 CPU-consuming processes
• get_top_processes_by_memory() - Fetches top 5 memory-consuming processes
• format_process_leaderboard() - Formats process info as markdown leaderboard

Updated Alert Functions:

• send_cpu_alert() - Now includes top CPU-consuming processes with their usage percentages
• send_ram_alert() - Now includes top memory-consuming processes with usage percentage and MB consumed

Example Output

When a CPU or RAM alert is triggered, the notification will now show:

Top Consuming Processes:
1. `process_name` (PID: 1234) - 45.2%
2. `another_process` (PID: 5678) - 23.1%
...

This gives system admins immediate visibility into which processes are causing the strain, making it much easier to diagnose and resolve issues quickly.

---

Branch: claude/issue-10-20260206-0417 |