molt-verify

Compares source and target databases for schema and row-level consistency after migration.

2|1|Updated Mar 11, 2026
One-click install
npx skills add https://github.com/cockroachdb/cursor-plugin --skill molt-verify-cockroachdb
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: molt-verify
Source: https://github.com/cockroachdb/cursor-plugin/tree/main/skills/cockroachdb-onboarding-and-migrations/molt-verify
Command: npx skills add https://github.com/cockroachdb/cursor-plugin --skill molt-verify-cockroachdb

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill includes references (resource) components.

What problem does it solve? After migrating data to CockroachDB with molt fetch, you need proof that the target database actually matches the source — missing rows, mismatched values, or schema drift can silently corrupt a cutover. This Skill guides you through running molt verify to confirm schema and data integrity before going live. ## Core Features & Use Cases - Schema and Row Verification: Compares table presence, columns, types, constraints, and primary keys, then iterates source rows in PK order to detect missing, extraneous, and mismatched rows. - Concurrency and Sharding Tuning: Parallelize verification across tables and within large tables using PK-range sharding, batch sizing, and rate limiting to control production impact. - Schema-Only CI Gates: Run fast DDL-only checks with --rows=false and --non-interactive for automated pipelines. - Use Case: After migrating a PostgreSQL database to CockroachDB, run a full verify with table filters and column exclusions, review per-shard mismatch reports, and fix discrepancies before cutover. ## Quick Start Ask the AI to run molt verify against your source PostgreSQL database and target CockroachDB cluster to confirm the migration copied all rows correctly.

Frequently Asked Questions about molt-verify

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I verify data after migrating to CockroachDB with molt?▼

Run molt verify with --source and --target connection strings after molt fetch completes. It compares schema first, then iterates source rows in primary key order, reporting missing, extraneous, and mismatched rows per shard.

How to run a schema-only check with molt verify?▼

Pass --rows=false to skip data comparison and only check table presence, columns, types, constraints, and primary keys. Combine with --non-interactive and --log-file stdout for CI pipeline gates.

Does molt verify support Oracle source databases?▼

Yes, but the binary must be built with CGO_ENABLED=1 and the cgo and source_all tags, with Oracle Instant Client in LD_LIBRARY_PATH. Selective row filtering via --filter-path is not supported for Oracle sources.

How do I speed up molt verify on large tables?▼

Use --concurrency-per-table to shard a table's primary key range across workers and increase --row-batch-size. Sharding supports INT, FLOAT, DECIMAL, and UUID primary keys; other types fall back to a single full scan.

Why does molt verify report missing or mismatched rows?▼

Missing rows exist on the source but not the target, extraneous rows exist only on the target, and mismatches mean values differ. Common causes include incomplete fetch runs, type conversion issues, or writes occurring between fetch and verify.

What are the limitations of molt verify row comparison?▼

Row verification requires primary keys on both source and target tables; tables without PKs are skipped. PostgreSQL partition tables are unsupported, and schema changes made after migration are not automatically reconciled.