mysql_audit_ai_agent

MySQL schema-change safety review for AI coding agents.

Before Claude Code, Cursor or another agent runs a migration, it sends the SQL to review_migration and gets back a verdict (safe / review / block) plus, for every risk, the line, what will happen in production, and a safe way to write it. The checks are deterministic rules written from DBA experience, not another LLM guess.

Status: early development (v0.1). All twelve planned rules are in.

Rules

Tables created earlier in the same script are treated as empty, so DDL on them is not flagged.

• schema: the current CREATE TABLE statements (SHOW CREATE TABLE).
• stats: approximate row counts, {"orders": 25000000}, from SELECT table_name, table_rows FROM information_schema.tables WHERE table_schema = DATABASE(). Tables of 1,000,000 rows or more count as large; under 10,000 as small.

Use with an AI agent (MCP)

Requires uv. The server runs locally over stdio; your SQL never leaves your machine.

Claude Code

claude mcp add mysql-audit -- uvx mysql-audit-ai-agent

Cursor, Windsurf, Claude Desktop and other MCP clients (mcp.json / claude_desktop_config.json):

{
  "mcpServers": {
    "mysql-audit": { "command": "uvx", "args": ["mysql-audit-ai-agent"] }
  }
}

Claude Desktop, one click: download mysql-audit.mcpb from the latest release and open it.

Tools: review_migration(sql, schema?, stats?), review_query(sql, schema?, stats?), review_code(code), explain_rule(rule_id), list_rules().

In our agent evaluation Claude Code called the tool on its own in 10 of 10 migration tasks, with no extra prompt.

Command line

pip install mysql-audit-ai-agent
mysql-audit migrations/20261027_add_index.sql --schema schema.sql
mysql-audit migration.sql --stats rows.json --format json
mysql-audit --code app/models/user.py      # SQL injection check

Exit code: 0 safe, 1 review, 2 block, so it can gate CI.

GitHub Action

Reviews the migrations each pull request adds or changes, annotates the lines in the diff, keeps one summary comment on the pull request up to date, and fails the check on block.

# .github/workflows/mysql-audit.yml
name: MySQL Audit
on: pull_request
permissions:
  contents: read
  pull-requests: write   # for the summary comment
jobs:
  audit:
    runs-on: ubuntu-latest
    steps:
      - uses: actions/checkout@v4
      - uses: jimmyjan0824/mysql_audit_ai_agent@v0.2.0
        with:
          paths: db/migrate/**/*.sql      # default: **/*.sql
          schema: db/schema.sql           # optional, makes checks exact
          # stats: db/table_rows.json     # optional row counts, enables large-table rules
          # code-paths: app/**/*.py       # optional SQL injection check
          # fail-on: review               # block (default), review or never

Development

pip install -e ".[dev]"
pytest

Every rule has positive and negative test cases under tests/.

tests/test_mysql_oracle.py checks the rules against a real server: each ALTER is run with ALGORITHM=INPLACE, LOCK=NONE, and MySQL's refusal (or not) must match the rule's verdict. CI runs it on MySQL 8.0 and 8.4. Locally:

docker run -d -e MYSQL_ROOT_PASSWORD=root -p 3306:3306 mysql:8.4
MYSQL_HOST=127.0.0.1 MYSQL_PASSWORD=root pytest tests/test_mysql_oracle.py

中文說明

給 AI 編碼代理用的 MySQL 變更安全審查。代理在執行 migration 前呼叫 review_migration， 取得 safe / review / block 判定，以及每個風險的行號、線上影響與安全寫法。 規則由 DBA 經驗寫成，結果可重複驗證。

License

Apache-2.0