Release Incident Playbook
Release Incident Playbook documents 84 types of production incidents caused by software releases, each with at least two public post-mortems, the test class that would have caught it, a graded risk card, and downloadable pre-release and first-30-minutes checklists.
Public examples only. Tools and providers listed by what their own documentation says, in alphabetical order, with no ratings.
Three entry questions
Why did tests not catch this?
Find the incident type, read its public examples, and see which test class catches it and what that class misses.
Browse incident typesWhat goes into the next regression scope?
Each incident has a pre-release checklist tied to test classes and a first-30-minutes checklist.
Browse checklistsIs this class of failure covered by a tool, a contractor, or neither?
Test class pages list the 89 tools and 59 providers whose own documentation names that class.
Browse test classes
Incident groups
Method
Incident types come from public post-mortems, status page records, regulator findings and maintainer-marked repository issues. Each report is opened and its quoted fact confirmed before it counts. Grades are computed by script.
Methodology · Download the data
| Page type | Published | Withheld |
|---|---|---|
| Incident pages | 58 | withheld 26 |
| Checklist pages | 58 | withheld 26 |
| Tool pages | 89 | withheld 14 |
| Provider pages | 59 | withheld 56 |
Questions
What is the Release Incident Playbook?
Release Incident Playbook documents 84 types of production incidents caused by software releases, each with at least two public post-mortems, the test class that would have caught it, a graded risk card, and downloadable pre-release and first-30-minutes checklists.
What counts as a release incident in this playbook?
A release incident is a production failure caused by a change: a deploy, a configuration edit, a migration, an upgrade or a certificate or dependency change. An incident type is published only when at least two organisations have an accepted, dated public report of it, with at least one report from the last 5 years.
What types of production incidents does the playbook cover?
The playbook defines 84 incident types in 10 groups: payments and billing, data and migrations, authentication and access, performance and capacity, configuration and feature flags, mobile release, third-party integrations, frontend web, infrastructure and deploy and observability and rollback. 58 are published.
What causes production incidents after a release?
Across 58 published incident types, the root-cause stage named most often is code (33 incident types), followed by release-process (27) and config (23).
Why did tests not catch the bug?
Each incident page states the limitation of every test class that catches it. Each test class also has a fixed list of what it cannot catch; for example, canary and progressive delivery checks cannot catch faults in traffic the canary never receives.
How is the playbook built?
Every incident page follows the same 13 sections. Evidence comes from public post-mortems, status page records, regulator findings and maintainer-marked repository issues. Likelihood and impact grades are computed by script from those reports.