Snapshot and diff committed database state before and after actions.
npx skills add https://github.com/kylesnowschwartz/ralph-ban --skill db-state-qa
Please help me install this Agent Skill. Skill: db-state-qa Source: https://github.com/kylesnowschwartz/ralph-ban/tree/main/skills/db-state-qa Command: npx skills add https://github.com/kylesnowschwartz/ralph-ban --skill db-state-qa
This Skill verifies that an action’s intended database side effects actually occurred by snapshotting committed DB state before and after, then producing a structural diff that matches (or contradicts) the spec.
Use it when you need to assert behaviors such as “a row was persisted,” “a table was not modified,” “a cached counter updated,” or “a migration added an index/constraint,” especially in scenarios driven by http-qa, cli-qa, or library-qa oracles.
Use db-state-qa to snapshot the database state before an action, re-run the action through the appropriate oracle surface, then snapshot again and diff the normalized results against the spec’s asserted side effects.
High-intent search queries and answers about installing and using this skill.
Verify database side effects by snapshotting committed DB state before and after an action, then diffing the resulting rows and schema information. This avoids false approvals caused by rollback transactions by checking live, committed rows directly.
Database state diffing captures row sets and schema information using stable ordering, then normalizes volatile fields like ids and timestamps to reduce noise. This deterministic structural diff reliably matches or contradicts your spec's asserted persistence behaviors.
Test Rails ActiveRecord schema changes using engine-specific catalogue introspection for PostgreSQL, SQLite, and MySQL. This approach avoids stale schema caches and WAL/MVCC snapshot pitfalls by querying the database catalogue directly rather than relying on cached schema data.
Database state verification supports waiting for quiescence so after-snapshots don't race queued jobs, after-commit callbacks, or replication lag. This async-aware validation ensures the snapshot captures all committed changes reliably before diffing.
You can use database snapshot diffing to assert the absence of modification by comparing pre-action and post-action snapshots. If the structural diff shows no changes for the target table, the verification reliably approves the spec's assertion.
Snapshot-based database side effect testing requires deterministic snapshots with stable ordering, volatile-field normalization, and quiescence-aware timing. Without proper engine-specific catalogue introspection and timing control, asynchronous jobs or MVCC snapshot pitfalls can cause unreliable verdicts.
rauell1Configure Prisma database connections and generate clients for multiple providers.
mkrlabsGenerate structured QA verdicts and metrics for validation workflows.
concept2cureAutomate browser-based QA with a persistent headless browsing daemon.
reliant-labsRun database-backed Go integration tests against a real Postgres instance.
