Complete AI Training
Sign inGet my AI kit

Your job's AI kit

Get your AI kit

Tell us who you are and what you do. We show you your kit right away and email you the link: skills, prompts, AI agents, MCP servers and courses for your job.

500+ jobs ready, and we make a kit for any other job. No payment needed to look.

Share

AI agent for site reliability engineers

Blockchain Event Indexer Health Agent

The indexer stays close to the chain head with no missing events

Blockchain Event Indexer Health Agent: what goes in, what the agent does and what you get

What it does

Apps that show blockchain data rely on an indexer that can lag, miss events during chain reorganizations or crash, and then users see wrong balances. Every few minutes this agent checks how far behind the indexer is, compares its event counts with the chain for recent blocks and looks for gaps. For known safe issues, such as a stalled connection, it applies the runbook action, like reconnecting, and checks whether the lag clears. If events are missing, it triggers a re-index of the affected block range and verifies that counts now match. If they still differ, it raises a ticket with the block range and the counts. You approve anything that changes stored data beyond a re-index. Edge case: a reorganization is handled by re-indexing the rolled-back blocks, never by deleting records blindly.

How it works

Follow the arrows from top to bottom. The orange dashed arrow is the loop: when a check fails, the agent goes back and tries again.

Start and resultWhat it doesA check on its own workWaits for your OKGoes back and retries
Yes, continueYes, continueApprovedNoNo 1 STARTS WHEN Scheduled indexer check 2 USES A TOOL Read indexer lag and recent event counts 3 USES A TOOL Compare counts with the chain for recent blocks 4 CHECKS THE RESULT Is the lag within limit and are counts matching? If not: apply a runbook action or trigger a re-index ofthe affected range. Back to step 2. 5 USES A TOOL Re-verify counts after the action 6 CHECKS THE RESULT Do the counts now match the chain? If not: raise a ticket with the block range and gap.Back to step 4. 7 YOU APPROVE Developer approves actions that rewrite stored data 8 RESULT Indexer health log
Read the steps as a list
  1. Scheduled indexer check
  2. Read indexer lag and recent event counts
  3. Compare counts with the chain for recent blocks
  4. Is the lag within limit and are counts matching?If not: apply a runbook action or trigger a re-index of the affected range. Back to step 2.
  5. Re-verify counts after the action
  6. Do the counts now match the chain?If not: raise a ticket with the block range and gap. Back to step 4.
  7. Developer approves actions that rewrite stored dataThe agent waits here for your OK.
  8. Indexer health log

How it decides

It acts automatically only on runbook items, and confirms health only when the lag is within limit and recent event counts match the chain.

  • Warn when lag exceeds the set number of blocks
  • Re-index only the affected block range
  • Handle reorganizations by re-indexing rolled-back blocks

Make it yours

Every agent is a starting point. You choose these settings for your own situation.

  • Lag limit in blocks
  • Check frequency
  • Runbook actions allowed
  • Block ranges or contracts to watch

What keeps you in control

It always asks you first

  • Rewriting or deleting stored event data

Hard limits

  • No data rewrite without approval
  • Read-only by default

It stops when

  • Done: indexer healthy and counts match
  • Stop: chain node unreachable

Set it up

We guide you through the set-up, step by step

Members get the full set-up guide for this agent. No technical skills needed: you copy, paste and upload.

10 minto set it up in your AI
5 AIsChatGPT, Claude, Copilot, Gemini, Grok
  • One set of instructions to paste into your AI, with the clicks for ChatGPT, Claude, Microsoft 365 Copilot, Gemini and Grok
  • The agent then walks you through connecting your own data, one source at a time
  • A downloadable copy with the flow chart, the rules and the full guide
Get access to this agent

An example run

What happensAt 14:20 on a Friday, the indexer for a wallet app at Tallgrass Labs was 180 blocks behind, over the 50-block limit. A reconnect cleared the lag, but the count check failed: blocks 44,210 to 44,260 had 12 fewer transfer events than the chain after a reorganization. The agent re-indexed that range and counts matched. The platform lead approved closing the alert.

More agents for site reliability engineers