OpenParable
Skip to content
← Back to skills

Schema Change

Development
schema-change

Schema Change is an Agent Skill for development published in the kagkarlsson/db-scheduler repository on GitHub: a SKILL.md instruction file an AI agent can load to do this task.

Use when changing the db-scheduler database schema (adding/removing/altering columns, indexes, or the scheduled_tasks table). Walks through the lock-step update across every supported DB dialect, the matching JdbcCustomization changes, and the documentation copies. Skip for code-only changes that don't touch table structure.

View on GitHub ↗
QUICK START

How to use this skill

Bring this guide into your coding agent with a prompt tailored to the tool you use.

  1. Open your project in Codex.
  2. Copy and paste the prompt below.
  3. Review files and risks before approving any action.
Prompt to paste
Inspect the Agent Skill at https://github.com/kagkarlsson/db-scheduler/blob/HEAD/.claude/skills/schema-change/SKILL.md and all supporting files. Report provenance, permissions, scripts, dependencies, and security risks. Propose an exact installation into .agents/skills/schema-change, but wait for my approval before writing files or running scripts.

Copying this prompt does not install or run the skill. Review third-party files before use.

Schema change checklist

The library supports six DB dialects and ships DDL for each. A schema change that misses one dialect ships a broken release. Treat this as a single atomic change.

1. DDL files (test resources)

Update all six in lock-step. Path: db-scheduler/src/test/resources/

  • hsql_tables.sql
  • postgresql_tables.sql
  • mysql_tables.sql
  • mariadb_tables.sql
  • mssql_tables.sql
  • oracle_tables.sql

Read all six first; column ordering and types differ per dialect (e.g. TIMESTAMP WITH TIME ZONE vs DATETIME2 vs TIMESTAMP(6) WITH TIME ZONE). Don't blindly copy from one dialect to another — preserve each dialect's existing type conventions.

There's also postgresql_custom_tablename.sql under src/test/resources/com/github/kagkarlsson/scheduler/ — check whether your change needs to be reflected there too (it tests configurable table name; usually yes if columns change).

2. JdbcCustomization classes

Path: db-scheduler/src/main/java/com/github/kagkarlsson/scheduler/jdbc/

If your change affects how rows are read/written (new column, type change, nullable→ not-null, etc.), update each relevant class:

  • DefaultJdbcCustomization.java — base behavior
  • PostgreSqlJdbcCustomization.java
  • OracleJdbcCustomization.java
  • MssqlJdbcCustomization.java
  • MySQLJdbcCustomization.java
  • MySQL8JdbcCustomization.java
  • MariaDBJdbcCustomization.java
  • AutodetectJdbcCustomization.java — only if dialect-detection logic changes
  • JdbcCustomization.java (interface) — if the contract itself changes

Also check JdbcTaskRepository.java and QueryBuilder.java for SQL that references the changed columns.

3. Documentation DDL copies

The README and/or docs/ may carry copies of the DDL for users. Grep for a column name from the table to find them:

grep -rn "scheduled_tasks" --include="*.md" .

Update any matching SQL blocks so users get the current schema.

4. Verify

mvn -pl db-scheduler test                                # PostgreSQL tests

Dialects other than PostgreSQL currently need to be tested on CI (need to run on amd64 arch).

5. Migration note for users

Schema changes are user-visible. Add a brief migration note to the changelog / release notes describing the required ALTER TABLE (per dialect if they differ).

6. Finalize

Run /finalize before committing.