🚨 ShardWorks Update: Monitoring & Alarm System Is Live

I am currently building ShardWorks to operate my Non-Battle Splinterlands operations and I try to build it publicly. For some friendly users, I can grant access and would be happy to receive feedback on functionality and usability. If you're interested, feel free to reach out in the comments.

---

🗺️ What Is ShardWorks?

ShardWorks is a hosted automation platform for Splinterlands and HIVE accounts. It handles repetitive on-chain operations on a
schedule — things like:

• Collecting SPS to a bank account
• Staking SPS and claiming staking rewards
• Claiming daily rewards
• Harvesting land and adding liquidity
• Buying season passes and distributing DEC

Once configured, it runs all of this automatically in the background. No browser tabs to keep open, no manual clicking.

You can do most work with your posting key. Some actions require an active key, but it's of course up to you to use those features!

---

🆕 What's New: Monitoring & Alarms

The latest update introduces a dedicated Monitoring tab with three major capabilities.

---

🩺 Job Health Dashboard

The Monitoring tab now gives you an at-a-glance view of every job you have running. Each job shows a coloured health badge:

Status	Meaning
🔴 ALERTING	One or more alarm rules are triggered
🟢 OK	All rules evaluate to normal
🟡 NO DATA	Not enough run history yet

At the top of the dashboard, summary counters show how many jobs are alerting, OK, or waiting for data — so you can instantly spot trouble
without scrolling through a list.

---

⚙️ Configurable Alarm Rules

The most powerful part of this update: you decide what counts as a problem. Each job can have one or more alarm rules attached to it. Three rule
types are available:

🔁 Consecutive Failures

"Alert me if this job fails 3 times in a row."

Triggers when the last N runs are all failures. Ideal for catching jobs that are stuck in a broken state.

Example: threshold=3, window=5 → alerts if the most recent 3 completed runs all failed.

📉 Failure Rate

"Alert me if more than 60% of the last 10 runs failed."

Triggers when the failure rate across a sliding window exceeds your threshold. Useful for jobs that are flaky but not entirely dead.

Example: threshold=0.6, window=10 → alerts if 6 or more of the last 10 runs failed.

⏰ Missed Schedule

"Alert me if this job hasn't run and is 2 hours overdue."

Triggers when a job is past its scheduled time by more than N hours. Catches situations where the worker itself may have a problem, or a job got
stuck mid-run.

Example: threshold=2.0 → alerts if the job is more than 2 hours late starting.

Rules are managed through a clean modal in the Monitoring tab — click the Rules button next to any job (a badge shows how many rules are already
configured), add or remove rules, and they take effect immediately.

---

📧 Email Notifications

Knowing there's a problem is great. Knowing immediately is better.

You can now enter a notification email address in the Monitoring tab. When an alarm triggers, ShardWorks sends you an email with the job name,
account, run ID, and the error message — everything you need to act fast.

Subject: [ShardWorks] Job failed: collect-sps

A scheduled job has failed.

Job: collect-sps
Account: myaccount
Started: 2026-03-09T10:30:00+00:00
Error: operation_failed: connection timeout

You can also enable resolve notifications — a second email when the alarm clears and the job is healthy again. That way you know the problem
fixed itself (or that your fix worked) without having to log back in.

---

📋 Alarm Event History

The Monitoring tab also shows the 50 most recent alarm events with timestamps — every time an alarm fired or resolved, you can see it. This
gives you a short audit trail to understand whether a problem was brief or persistent.

---

🔐 Access Control

The Monitoring feature is gated behind an allowlist on the server side, so it can be enabled for specific Hive accounts only. If your account
isn't on the list, the Monitoring tab is hidden entirely and all monitoring API endpoints return 403. This keeps the feature restricted to
approved users while the platform is in early access.

---

💡 Why This Matters

Before this update, the only way to know a job was failing was to:

1. Log into the dashboard
2. Navigate to the Jobs tab
3. Select each job individually
4. Read through the run history

Now, the Monitoring tab gives you a single screen with everything at a glance, plus email pushes the alerts to you so you don't have to go
looking.

For anyone managing multiple accounts and operations, this is the difference between catching a failure in minutes vs. discovering it the next
day when your SPS didn't arrive.

---

🛠️ Coming Next

The foundation is now in place for more advanced monitoring capabilities. Potential directions include more rule types, webhook notifications
(Discord, Telegram), and longer event history windows.

If you're using ShardWorks and have feature requests, drop them in the comments. 👇