TiDB Schema Change Audit for Safer DDL

DeltaScope is an offline-first TiDB schema change audit tool that analyzes DDL statements before they reach your TiDB cluster. It catches destructive ALTER TABLE operations, missing defaults, incompatible type changes, and dozens of other risky patterns in CI pipelines, developer workflows, and AI agent sessions.

How it works

Pass one or more DDL statements to DeltaScope with --dialect tidb. It parses the SQL using TiDB-compatible syntax, evaluates every applicable rule, and returns a verdict with blockers, warnings, and notices. No database connection is needed for offline audit.

$ deltascope audit --dialect tidb --sql "ALTER TABLE users DROP COLUMN email;"

Verdict: reject
Statements: 1
Blockers:   1
Warnings:   0
Notices:    0

Statement 1: ALTER TABLE
- [blocker] ddl.alter.drop.column: dropping column "email" is a destructive change

Common TiDB DDL checks

• blocker Dropping columns or tables
• blocker Changing column types incompatibly
• warning Missing column defaults
• warning Missing column comments
• notice Redundant index additions

TiDB schema change review with metadata

Connect to a live TiDB instance to enrich findings with table statistics:

deltascope audit \
  --dialect tidb \
  --sql "alter table users add column email varchar(255)" \
  --host 127.0.0.1 --port 4000 --user root --ask-password --schema app \
  --metadata-connect-timeout 5s

Use in CI

# GitHub Actions example
- name: Audit TiDB schema changes
  run: |
    deltascope audit \
      --dialect tidb \
      --file ./migrations/tidb_schema_change.sql \
      --format github-actions \
      --fail-on blocker

DeltaScope also supports --format sarif for GitHub Code Scanning and --format gitlab-codequality for GitLab CI.

Multiple integration paths

Use the same audit engine from the CLI, in CI, via HTTP service, or through the MCP server for AI agents. Every path uses the same rules and the same policy configuration.

Frequently asked questions