Skip to content
Skills Library
Programming · Review & audit

database-failure-modes

/database-failure-modes

Failure-mode checklists for database implementation work, split by domain. Use when implementing or reviewing changes that touch schema migrations and deploy ordering, locks and online DDL, indexes and query plans, constraints and integrity, column types and semantics, backfills and large tables, transactions and connection poolers, ORM schema artifacts and drift, or roles, grants, and row-level security. Read only the reference files whose triggers match the change.

Licence
MIT
Last repository change
10.11.2026

Assessment and security

90/ 100Class: good
Format25 / 25
Frontmatter with name and description, valid name, description up to 1,024 characters, name matching the folder.
Discoverability25 / 25
Description explains when to use the skill, has a reasonable length and uses third person.
Content25 / 35
At least 150 words and at most 500 lines, headings, step lists, examples or code.
Hygiene15 / 15
No placeholders such as TODO, no Windows paths, with a licence.
Riskno signals detected in the scan
Files in this package
11
Scripts
0
HTML comments
0
URLs in SKILL.md
0
Deprecation

No signals detected. Repository changed on 10.11.2026, without references to retired models or APIs.

Licence

MIT, open: use and modify while retaining author attribution.

What it does and when to use it

Failure-mode checklists for database implementation work, split by domain. Use when implementing or reviewing changes that touch schema migrations and deploy ordering, locks and online DDL, indexes and query plans, constraints and integrity, column types and semantics, backfills and large tables, transactions and connection poolers, ORM schema artifacts and drift, or roles, grants, and row-level security. Read only the reference files whose triggers match the change.

Original author description

Failure-mode checklists for database implementation work, split by domain. Use when implementing or reviewing changes that touch schema migrations and deploy ordering, locks and online DDL, indexes and query plans, constraints and integrity, column types and semantics, backfills and large tables, transactions and connection poolers, ORM schema artifacts and drift, or roles, grants, and row-level security. Read only the reference files whose triggers match the change.

How it works

  1. Trigger table
  2. How each reference is structured
Original SKILL.md headings · 35 lines, 434 words
SKILL.md preview
---
name: database-failure-modes
description: >-
  Failure-mode checklists for database implementation work, split by domain.
  Use when implementing or reviewing changes that touch schema migrations and deploy ordering,
  locks and online DDL, indexes and query plans, constraints and integrity, column types and
  semantics, backfills and large tables, transactions and connection poolers, ORM schema
  artifacts and drift, or roles, grants, and row-level security.
  Read only the reference files whose triggers match the change.
---

# Database failure modes

Checklists of the ways database changes go wrong in production, one reference file per domain.
This skill is a router: match the change against the trigger table, read only the files that fire, and treat every checklist item in them as a gate.
An unresolved item blocks `done`; when the brief itself forces the failure mode, escalate with `needs-decision` instead of shipping it.

## Trigger table

| The brief or diff touches... | Read |
|---|---|
| New migrations, renames or drops, deploy sequencing, anything applied in more than one step | [references/migrations-and-deploy-ordering.md](references/migrations-and-deploy-ordering.md) |
| Any DDL against an existing table: index builds, column or default changes, type changes, constraint adds | [references/locks-and-online-ddl.md](references/locks-and-online-ddl.md) |
| New or changed indexes, query changes, plan evidence, anything labeled "slow" | [references/indexes-and-query-plans.md](references/indexes-and-query-plans.md) |
| Foreign keys, unique constraints, NOT NULL, checks, cascade rules, invariants enforced in app code | [references/constraints-and-integrity.md](references/constraints-and-integrity.md) |
| Column types chosen or changed, timestamps, money, text and charsets, IDs and sequences | [references/data-types-and-semantics.md](references/data-types-and-semantics.md) |
| Backfills, mass UPDATE or DELETE, moving data between columns or tables | [references/backfills-and-large-tables.md](references/backfills-and-large-tables.md) |
| Transaction wrapping of migrations, isolation assumptions, locking reads, connection poolers | [references/transactions-and-concurrency.md](references/transactions-and-concurrency.md) |
| ORM schema files, generated clients or types, migration state, multiple heads, drift | [references/schema-drift-and-artifacts.md](references/schema-drift-and-artifacts.md) |
| Roles, grants, object ownership, row-level security policies, stored routine bodies, the accounts application or agent code connects as | [references/privileges-and-row-security.md](references/privileges-and-row-security.md) |
| Migration and backfill failure legibility, progress and resume signals, slow-query and lock visibility | [references/failure-visibility.md](references/failure-visibility.md) |

Most real changes fire two or three rows (a typical schema-change brief fires at least migrations-and-deploy-ordering, locks-and-online-ddl, and schema-drift-and-artifacts).
Read all of them; skip the rest.

## How each reference is structured

- **Failure modes to rule out**: concrete ways the change breaks in production, each with a `Check:` you can actually perform against the diff.
- **Escalation triggers**: conditions that are decisions, not implementation details; report `needs-decision` with a recommendation.
- **What good looks like**: the positive pattern, for calibration.

The checks are engine-agnostic on purpose: engine behavior is named only where the failure mode is engine-defined (transactional DDL, `CONCURRENTLY` semantics); ORM- and tool-specific guidance belongs to the stack skills the caller has installed, not here.
The exhaustive per-recommendation depth for security surfaces (database privileges, row-level security, injection in stored routines, vector store access) lives in `owasp`, which these checklists summarize rather than replace.
Transaction semantics and data safety inside application code belong to the backend seat's data-safety reference; lakehouse and analytical tables belong to the data seat: these files own OLTP DDL mechanics, migration ordering, lock behavior, and plan evidence.
Files in this package (11)
  • SKILL.md 4 KB
  • references/backfills-and-large-tables.md 3 KB
  • references/constraints-and-integrity.md 3 KB
  • references/data-types-and-semantics.md 3 KB
  • references/failure-visibility.md 2 KB
  • references/indexes-and-query-plans.md 4 KB
  • references/locks-and-online-ddl.md 4 KB
  • references/migrations-and-deploy-ordering.md 3 KB
  • references/privileges-and-row-security.md 4 KB
  • references/schema-drift-and-artifacts.md 3 KB
  • references/transactions-and-concurrency.md 3 KB

Versions

The database has no other skills named database-failure-modes.

Database snapshot: 13.11.2026 · package version b3989b01dd11 (SHA-256 digest of the entire package)

Downloads in the last 30 days: 0 · ranking

Report a problem with this skill: harmful content, incorrect descriptions or an author’s removal request.

Similar skills

  • 90frontend-failure-modes
    Failure-mode checklists for frontend implementation work, split by domain. Use when implementing or reviewing frontend changes that touch component state and rendering, data fetching and client caches, accessibility, performance, forms, error handling and resilience,…
  • 90backend-failure-modes
    Failure-mode checklists for backend implementation work, split by domain. Use when implementing or reviewing backend changes that touch API contracts, schema migrations, queues and background jobs, cross-service calls, authn/authz, caching, telemetry, query performance, or…
  • 100database-design
    Database decisions - DB and ORM, schema, indexes, queries, migrations. Use when designing or changing a data model.
  • 90database-design-review
    Reviews schema and migration changes for integrity and safety. Use when a request changes tables, constraints, nullability, data movement or migration order, even in a small code change.
  • 90database-design
    Picks a database and ORM, designs schemas and indexes, plans safe migrations. Use when designing a schema, adding indexes or planning a migration.
  • 100database-engineering
    End-to-end database engineering workflow for designing relational data models, writing schemas and migrations, reviewing constraints, creating indexes, optimizing queries, and hardening persistence behavior. Use when the user asks to design a database, create or change tables,…

Feedback and reports

Your message goes to Bartosz Sokoliński. An email address is needed only for a reply. See the privacy policy.