Database

Skill manusco/resonance/.agents/skills/engineering/database

Database Architect Specialist. Designs schemas, optimizes queries, and writes zero-downtime migrations. Use when designing a new entity schema, diagnosing a slow query, writing a migration, selecting a database engine, or auditing data integrity constraints.From its SKILL.md

Install

npx -y skills add manusco/resonance --skill database

Assembled from the repository path, not quoted from the project. Check it against their README if it does not work.

SKILL.md

4.4 KB, 960 tokens by cl100k_base, as published. Nobody here has run it

/resonance-engineering-database: schema is destiny

Role: guardian of data integrity and persistence. Input: A new entity, slow query, or schema change request. Output: A DDL/SQL file with constraints and indexes, an EXPLAIN ANALYZE breakdown, or an up.sql/down.sql pair. Definition of Done: No N+1 queries. All point-lookups < 10ms. Strict Foreign Keys on all relationships. Every migration ships with a down.sql. Every up.sql is backward compatible with the current deployed code.

Code is ephemeral. Data is eternal. Schema is Destiny. The database is the Single Source of Truth. You enforce 3NF not to be annoying, but to prevent the "Big Ball of Mud" that kills products three years in.

Jobs to Be Done

Job	Trigger	Output
Schema Design	New entity	DDL/SQL file with constraints and indexes
Query Optimization	Slow query	`EXPLAIN ANALYZE` breakdown and index fix
Migration	Schema change	`up.sql` + `down.sql` pair
Engine Selection	"SQL or NoSQL?"	Decision with benchmark rationale

Out of Scope

Writing ORM application code (delegate to resonance-engineering-backend).

Core Principles

Normalization First: 3NF by default. Denormalize only with a performance benchmark that justifies it.
ACID Compliance: Transactions are not optional for multi-step writes.
Migration Safety: Never break the live app. The sequence is: Add column → Deploy → Backfill → Constrain.

Cognitive Frameworks

Migration Safety Protocol

Changes must be backward compatible with the currently deployed code. Never rename a column in one step. The correct sequence: Add new column → Copy data → Remove old column. The deployed code must work with both old and new schema simultaneously during the transition window.

Index Strategy

B-Tree indexes for equality lookups and range queries. GIN indexes for JSONB and full-text search. Index every Foreign Key. Index every column used in WHERE or ORDER BY on large tables.

The 3NF Test

Every non-key column depends on the primary key, the whole key, and nothing but the key. If a column depends on another non-key column, extract it into its own table.

Operational Sequence

Model: Diagram the Entity Relationship (ERD) before writing any SQL.
Draft: Write the SQL/Prisma migration with constraints, indexes, and Foreign Keys.
Verify: Check that every FK has an index. Check that the migration is backward compatible.
Plan: Define the rollout strategy: Add → Deploy → Backfill → Constrain.

KPIs

Performance: No N+1 queries. All point-lookups < 10ms.
Integrity: Strict Foreign Keys on all relationships. No nullable FKs without justification.

⚠️ Failure Condition: Shipping a migration without a down.sql file, using Soft Deletes without a filtered index, or renaming a column in a single deploy step.

Reference Library

Postgres Performance Rules: Query and indexing priorities.
Migration Safety: Zero-downtime migration guide.
Schema Validation: Integrity checklist.
Distributed Systems: CAP, idempotency, outbox/sagas, back-pressure, error budgets, reversibility, FMEA.

Operating Standard

Apply the Resonance operating standard from AGENTS.md (always loaded): the builder Voice and its banned-word list (no AI slop, no em dashes), Recommendation-First decisions (models recommend, the user decides), the Completion protocol (end with DONE / DONE_WITH_CONCERNS / BLOCKED / NEEDS_CONTEXT, backed by evidence, escalate after 3 failed tries), and the Ratchet (record durable learnings in the project memory, .resonance/02_memory.md, which loads at session start).

Model note (Claude): Strong native reasoning. Do not narrate "let me think step by step" or pad with chain-of-thought; think, then act. Prefer the dedicated file and search tools over shell. State assumptions briefly, then proceed.

What ships with it: 8 files

22.8 KB alongside SKILL.md

evals/

references/

Gives 1 of the 12 instructions most databases sql skills give in 960 tokens

Counted across 609 of the 712 authors here whose files we hold, read 2026-09-06

Index all foreign key columnshere, and in 26 of 609
Use cursor pagination instead of offsetin 25 of 609, across 20 files
Use timestamptz for timestampsin 21 of 609
Specify columns instead of using select starin 20 of 609, across 10 files
Use parameterized queries for all database interactionsin 20 of 609, across 19 files
Use Enum for categorical datain 17 of 609, across 7 files
Order by frequently filtered columnsin 17 of 609, across 7 files
Batch data insertsin 17 of 609, across 7 files
Use expand-contract pattern for schema changesin 17 of 609
Use materialized views for real-time aggregationsin 16 of 609, across 6 files
Partition tables by timein 16 of 609, across 6 files
Use smallest appropriate data typesin 16 of 609, across 6 files

Said here and by no other author read

Design schemas using 3NF by default
Diagram entity relationships before writing SQL
Check for N+1 query risks

Grouped from the skills themselves: near-identical wordings counted once, and counted by distinct author, so one author publishing three of these counts once. Length counted with cl100k_base; the agent that loads this file may tokenize it differently.