You fix it yourself, with the failure modes in hand
- Full runbook, version-locked
- Detection command and expected output
- 5+ failure modes with verbatim errors
- Evidence appendix, re-runnable
- 7 days of questions (5)
- Refund within 7 days
One runbook costs what an hour of a consultant costs, and answers the question that hour would answer. The call tier exists because some environments really are different; the done-for-you tier exists because some boxes should not be touched by their owner at 1 a.m.
All prices are in USD and charged once. Payments and tax are handled by Paddle as merchant of record.
You fix it yourself, with the failure modes in hand
Your environment does not match the stated one, and you want a person on it
You want it working, not explained
All tiers are one-time USD purchases. Email [email protected] if you need an invoice.
The difference is not the amount of text. It is how much of your specific environment gets looked at.
| Included | One runbook | Runbook + call | Done for you |
|---|---|---|---|
| Version-locked runbook | yes | yes | — |
| Detection command | yes | yes | yes |
| Failure modes + verbatim errors | yes | yes | yes |
| Questions (7 days, 5 of them) | yes | yes | — |
| Review of your config | — | 60 min | full |
| Someone else does the change | — | — | yes |
| Refund | 7 days | until 24h before | per written scope |
Because it is not the text that costs $39, it is the verification work. The runbook states the version it targets, gives you a command to confirm you are affected, and lists the errors that only show up in logs. Most of that never makes it into free content, and none of it can be inferred from a model's training data after a recent change.
You can, and for anything older than a few months you probably should. The problem is the window that matters: right after a breaking change, the model answers about the previous behaviour with full confidence. That window is exactly what these runbooks cover.
The runbook lists its target environment on the first screen. If yours differs materially, ask first or take the $149 tier — do not buy the $39 one and hope. If the runbook fails anyway, the refund is automatic.
Seven days from purchase, full amount, no reason needed. Reply to the Paddle receipt or email us; Paddle processes the refund as merchant of record. The full policy is on the refund page.
No, and that is deliberate. Each runbook carries a review date and is re-checked quarterly; if it can no longer be checked it is delisted rather than silently going stale. What you buy is the version that matches its review date — see the changelog.
Runbooks are written by an AI from requests. Submit the platform change, the versions affected and public evidence; accepted requests become the next runbook.
The detection command is free. If it says you are not affected, you just saved $39.