Dev10x-db

Plan read-only SQL queries against project schemas for Claude Code.

Updated Aug 23, 2026
One-click install
npx skills add https://github.com/Brave-Labs/dev10x --skill dev10x-db-brave-labs
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: Dev10x-db
Source: https://github.com/Brave-Labs/dev10x/tree/main/codex-skills/dev10x-db
Command: npx skills add https://github.com/Brave-Labs/dev10x --skill dev10x-db-brave-labs

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

Database queries often risk unsafe writes and schema drift. This Skill enforces a schema-first, read-only approach for Claude Code to prevent production data changes and improve query correctness.

Core Features & Use Cases

  • Schema-driven query planning that loads context from memory files and project-specific skills.
  • Read-only enforcement: only SELECT queries are allowed; non-SELECT queries are emitted with guidance to run manually.
  • Engine-agnostic coordination: delegates execution to engine wrappers (e.g., Dev10x:db-psql) for the target database.
  • Schema validation against information_schema or project schema memory files to prevent guesses.

Quick Start

Provide a read-only SQL query plan against your project schema using the engine wrapper appropriate for your database.

Frequently Asked Questions about Dev10x-db

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

FAQPage Schema
How do I enforce read-only SQL queries in Claude Code?

To enforce read-only SQL queries in Claude Code, you need a schema-first planning layer that restricts operations to SELECT statements. This Skill validates queries against project schema memory files and blocks raw DDL or DML from executing automatically.

What is schema-first database query planning and how does it prevent schema drift?

Schema-first database query planning loads context from memory files and project-specific skills to validate table structures before generating SQL. By validating against information_schema, it prevents schema drift and avoids guesswork in query construction.

Can I use this database query planner with different SQL database engines?

Yes, you can use this query planner with different SQL database engines because it provides engine-agnostic coordination. It delegates execution to engine-specific wrappers, such as a PostgreSQL wrapper, after loading the target database DSN configuration.

How do I plan safe database queries using information_schema in Claude?

You plan safe database queries by loading schema context from information_schema or project schema memory files before writing SQL. The planner enforces SELECT-only operations, applies explicit timeouts, and routes execution through engine-specific wrappers for safety.

What happens when a non-SELECT query is generated during database query planning?

When a non-SELECT query is generated, the read-only enforcement mechanism blocks automatic execution. The system emits the non-SELECT query with guidance for you to run it manually, preventing unintended production data changes.

Do I need project-specific aliases and DSN configuration to run read-only database queries?

Yes, you need project-specific aliases and DSN configuration loaded before execution. These define the scope and connection target for your database, ensuring the schema-first query planning operates within the correct project boundaries.