db-mysql

Query MySQL / MariaDB databases declared in .env. Connections follow the same numbered pattern as SOCIAL_YOUTUBE_N_* / SOCIAL_INSTAGRAM_N_* — one block per database, labelled for humans, picked by label or index at call time.

Setup — one-time, per database

Add a block to .env (gitignored). Increment the index per connection:

# ── MySQL: orders-dev ────────────────────────────────
DB_MYSQL_1_LABEL=orders-dev
DB_MYSQL_1_HOST=mysql.dev.internal
DB_MYSQL_1_PORT=3306
DB_MYSQL_1_DATABASE=orders
DB_MYSQL_1_USER=agent_readonly
DB_MYSQL_1_PASSWORD=...                # raw; .env is gitignored
# DB_MYSQL_1_SSL_CA_PATH=/path/ca.pem  # optional, enables TLS verification
# DB_MYSQL_1_ALLOW_WRITE=false         # default false
# DB_MYSQL_1_QUERY_TIMEOUT=30          # seconds, default 30
# DB_MYSQL_1_MAX_ROWS=1000             # default 1000

Alternative — full DSN instead of components (DSN wins when both are set):

DB_MYSQL_2_LABEL=analytics-prod
DB_MYSQL_2_DSN=mysql://agent_ro:***@analytics.prod.internal:3306/analytics

LABEL is always required — it's how agents pick the connection.

Usage

All commands output a single JSON line on stdout (safe to pipe). Errors go to stderr as JSON and exit non-zero.

List configured connections

python3 .claude/skills/db-mysql/scripts/db_client.py accounts

Health-check a connection

python3 .claude/skills/db-mysql/scripts/db_client.py test orders-dev

Run a read-only query

python3 .claude/skills/db-mysql/scripts/db_client.py query orders-dev \
  "SELECT count(*) FROM customers WHERE created_at > now() - interval 7 day"

Explore schema

# List all tables (excludes mysql/information_schema/performance_schema/sys)
python3 .claude/skills/db-mysql/scripts/db_client.py tables orders-dev

# Describe a table
python3 .claude/skills/db-mysql/scripts/db_client.py describe orders-dev customers
# Schema-qualified:
python3 .claude/skills/db-mysql/scripts/db_client.py describe orders-dev analytics.events

Output shape

Successful query:

{
  "ok": true,
  "query_id": "uuid-v4",
  "label": "orders-dev",
  "columns": ["id", "email", "created_at"],
  "rows": [[1, "a@b.com", "2026-04-22 12:00:00"]],
  "row_count": 1,
  "truncated": false,
  "full_result_path": null,
  "execution_time_ms": 12
}

When rows exceed MAX_ROWS, truncated is true and full_result_path points to a CSV in ADWs/logs/db-queries/<query_id>.csv — the agent can read that file directly instead of re-running.

Error:

{"ok": false, "error_code": "write_blocked", "error": "Write query blocked — connection 'orders-dev' has ALLOW_WRITE=false. ...", "label": "orders-dev"}

Error codes:

• no_connections — no DB_MYSQL_N_* blocks in .env
• not_found — label/index doesn't match any block
• ambiguous — multiple blocks share the same label (use index instead)
• config_error — block present but required field missing (typically LABEL)
• driver_missing — pymysql not installed
• connection_failed — network, auth, TLS, or MAX_EXECUTION_TIME tripped
• write_blocked — write verb detected without ALLOW_WRITE=true
• multi_statement — more than one statement in a single call
• usage — wrong CLI args

Guardrails

• Write verbs (DELETE | UPDATE | INSERT | REPLACE | TRUNCATE | DROP | ALTER | CREATE | GRANT | REVOKE | RENAME | LOAD | CALL | HANDLER) — refused unless DB_MYSQL_N_ALLOW_WRITE=true.
• Multi-statement queries — refused in v1. Supports --, #, and /* */ comment styles.
• Query timeout — sets SESSION MAX_EXECUTION_TIME = <QUERY_TIMEOUT>s on the session (MySQL 5.7.4+; MariaDB honors max_statement_time but the SET statement is ignored gracefully on older versions).
• Result size — fetchmany(MAX_ROWS + 1) to detect truncation; full result streamed to CSV when truncated so the agent context never holds >1000 rows.
• utf8mb4 charset enforced on the connection to avoid mojibake on modern collations.

Workflow

1. If the user doesn't specify a label, run accounts to see what's configured and pick the one that matches their intent. If ambiguous, ask.
2. Write the smallest SQL that answers the question — prefer aggregates, LIMIT, and EXPLAIN before dumping rows.
3. Run via query. Inspect the result.
4. For performance-sensitive queries, load the deep-dive references below.

Dependencies

• Python 3.10+
• pymysql — not pre-installed; uv pip install pymysql on first use. Pure-Python driver, no system libs required.

Deep-dive references

These load the PlanetScale database-skills repo verbatim — same content the upstream authors ship for their own tooling.

• EXPLAIN analysis
• Composite indexes
• Covering indexes
• Fulltext indexes
• Index maintenance
• Primary keys
• Data types
• Character sets
• Isolation levels
• Deadlocks
• Row locking gotchas
• Online DDL
• Partitioning
• Replication lag
• Connection management
• N+1 queries
• Query optimization pitfalls
• JSON column patterns

Upstream: planetscale/database-skills (MIT). Credit to PlanetScale for the reference content.