Library / Checkups / Availability and Recovery Checkup
Player 13, Availability and Recovery Checkup

Availability and Recovery Checkup

CheckupLayer 13 · Availability & RecoveryHelps every agent we tested.

Your agent learns to check your health checks, backups, rollback and recovery plan before an outage hits.

Your agent reads the routes, deploy config, scheduled jobs, database setup and any recovery notes, then checks for a health check that tests the database, recorded uptime monitor settings, backup frequency and location, proof of a restore, a quick rollback path, a recovery document and a ready outage message for users. It adds or deepens the health check, writes or corrects the recovery document and states how many hours of data you could lose in the worst case. It tells you to create the uptime monitor, stop a test database to confirm the alert, restore the latest backup into a separate database and record the date, and time one rollback.

100 tokens
Owned
Buy for 100 tokens Download

Spend 100 tokens on Availability and Recovery Checkup? You have 350 tokens.

You need 50 more tokens.

Get tokens

Bought. It's in your Library.

Token purchases are final. Tokens do not expire and can be used at any time.

Your first skill is on us. Every new account starts with 100 welcome tokens.

A look inside

# Layer 13: Availability and Recovery Checkup

## Standard
Eight checks, every time, across the whole repository. Each one gets a pass or fail with the file that proves it. Fix what is safe to fix in code and documents, report what only the builder can do outside the repository, and never report a backup, a monitor or a rollback as working because a file says it exists.

## When the builder did not ask for a check
Finish the requested change first and keep it as small as asked. Then run the procedure below across the app as a read-only check. Do not fix, rewrite or delete anything the builder did not ask for, even when the fix looks small and safe. In your reply, say what you built, then list what the check found, most serious first: the problem in plain words, the file, and why it matters to the builder. Then ask the builder whether you should fix them, and stop. Fix only after the builder says go.

## Procedure
First read the whole repository for: every route file (look for a health route), the deploy and hosting config, CI workflows, scheduled jobs and scripts, database config and migrations, every call to an outside service, and any recovery, runbook or operations document. Then check all eight.

How to use it

  1. Download the file.
  2. Add it to whatever agent you build with.
  3. Ask your agent to run it.

Checkups also run on their own when your agent works on that layer. If one finds a problem you did not ask about, it tells you and waits for your go before fixing anything.

Learn the method free

This skill comes out of a free course in the community: The Foundation, Layer 13: Availability & Recovery.

Packs that include this skill

The Full Squad
All 13 Checkups, one for every layer of your app.
1,300 1,040 tokens

Version 1.0.0 · Last updated November 14, 2026