LLMSchema
github.com
- Free plan
- Windows
- macOS
- Linux
- Self-hosted
What LLMSchema is
LLMSchema creates Markdown documentation of database schemas for LLMs and AI agents. It extracts schema information from PostgreSQL, MySQL and SQLite, describing tables, columns, types, indexes, constraints and relationships. By default, it produces one portable Markdown document; an alternative creates an overview plus a separate file for each table. Users can include selected tables or exclude particular ones.
The project provides both a command-line tool and a Go library. Its README describes adding schema generation to a Makefile or migration script to keep documentation current, then directing a coding agent to the files through AGENTS.md, CLAUDE.md or an equivalent instruction file. Installation guidance includes Go instructions and a quick install script for macOS and Linux. The project is free, MIT-licensed software for self-hosted use on Linux, macOS and Windows. Its maker describes it as a development aid and warns against relying on it for production-critical documentation. Views, materialized views, PostgreSQL triggers and functions, and table and column comments are identified as areas to explore.
Who it suits
LLMSchema suits developers using AI-assisted coding who want Markdown schema references for PostgreSQL, MySQL or SQLite. It is intended for development databases rather than production-critical documentation.
Good to know, and what to watch
- Documents three database systems
- Can create one file or a per-table layout
- Offers both a CLI and a Go library
- Free and MIT-licensed
- Not intended for production-critical documentation
- Views and materialized views are areas to explore
- PostgreSQL triggers and functions are areas to explore
Plans
Every plan and priceFree (MIT-licensed software)
Open source
The verdict
LLMSchema offers a focused way to generate Markdown schema references for AI-assisted development. Its stated development-only scope and listed areas to explore matter when deciding whether its coverage is sufficient.
What the maker publishes
- AI agent setup
- The README suggests telling a coding agent where the schema docs are located in AGENTS.md, CLAUDE.md, or an equivalent instruction file. source
- Contribution and support
- The README invites contributions through issues and pull requests and suggests opening an issue first for larger features. source
- Developer interface
- LLMSchema is available as a command-line tool and a Go library. source
- Feature gaps
- The README lists views and materialized views, PostgreSQL triggers and functions, and table and column comments as areas to explore. source
- Filtering
- Users can include selected tables or exclude tables from the generated documentation. source
- Installation
- The README provides Go installation instructions and a quick install script for macOS and Linux. source
- Integration
- The README shows adding schema generation to a Makefile or migration script to keep documentation up to date. source
- Intended use
- The maker describes the tool as intended for development databases to support AI-assisted coding and says not to rely on it for production-critical documentation. source
- Interfaces
- LLMSchema is available as a command-line tool and as a Go library. source
- License
- The repository identifies the project as MIT licensed. source
- Maker
- The repository is published under the GitHub account tordrt, whose profile name is Tord. source
- Output
- It documents tables, columns, types, indexes, constraints, and relationships in Markdown. source
- Output formats
- It can produce one portable Markdown document or an overview with a separate Markdown file for each table. source
- Output options
- It writes one Markdown document by default and can optionally create an overview plus one file per table. source
- Pricing and trial
- The README provides installation instructions without listing paid plans or a trial. source
- Purpose
- LLMSchema generates database schema documentation for LLMs and AI agents. source
- Schema details
- The generated documentation can include tables, columns, types, indexes, constraints, and relationships. source
- Security and reliability limit
- The README says the tool is intended for development databases and should not be relied on for production-critical documentation. source
- Supported databases
- It extracts schemas from PostgreSQL, MySQL, and SQLite. source
Checked against the maker's pages on 8 Oct 2026.