Structure and ER map
Tables, columns, types, comments, composite and self-referencing FKs.
Learn more →Explore the database you already have
Big database.
Clear task.
SchemaMap is a visual explorer for existing database schemas. Explore tables, foreign keys and JOINs, collect task context and preserve it for your team and connected AI assistants.
Even a familiar system sometimes needs a reminder.
Full Hotflix schema · demo dataTask sheet · only what matters
Conceptual illustration · not a UI screenshot
EXISTING DATABASE → ONE TABLE → A CLEAR MAPEXPLORE THE EXISTING STRUCTURE ↗
A visual memory for your team
Recall how a familiar system fits together.
Start with one table; relationships bring context back.
●HOTFLIX / EXPLORATION
Right, payments connect to subscriptions
◈ subscriptionsA familiar table is selected● SYNTHETIC SCHEMA
“Now I remember how these connect.”
01 / OPEN02 / FIND RELATIONSHIPS03 / REBUILD CONTEXT
A conceptual story, not a recording of the interface.
02 / Exploration
Start with one table.
Related tables appear as ghosts. Add the ones you need with one click.
Concept: click a ghost table to add it to the sheet.
One table on the sheet
Ghosts add tables to the sheet. The depth switches above only change visibility.
03 / Working context
One database. Different tasks. Separate sheets.
Each task preserves an investigation the team can return to.
VIEW · PAYMENT
▦ users◈ id
▦ subscriptions◈ id · user_id
▦ plans◈ id
SYNTHETIC DATA
subscriptions.status = active●
SchemaTaskViewSheet
Interactive concept: independent sheet layouts. Not a product interface recording.
04 / MCP · REST API
AI prepares the map. You explore.
An agent finds tables and JOINs in code; you open the prepared task sheet.
01 / AGENT
Code
SELECT * FROM users JOIN subscriptions ON subscriptions.user_id = users.id→
02 / RELATIONSHIPS
JOIN
FK ────────── structure JOIN ────────── from code→
03 / SCHEMAMAP
Prepared sheet
Synthetic Hotflix SQL. A connected agent performs the analysis; access depends on the client and permissions.
Tools for database exploration
JOINs and SQL VIEWs
Dynamic JOIN overlay, join types, SQL/stack traces as evidence and VIEW dependencies. SQL is not executed.
Learn more →Navigation and ghosts
Add neighbors step by step; All / 1 step / 2 steps; hiding is separate from removal.
Learn more →Tasks, views and sheets
Independent sheets, duplicates, plan-checked moves, history and revision protection.
Learn more →Planned changes
Future tables, columns, relationships and indexes across a task; no migration execution.
Learn more →Team access
Owner/editor/viewer areas, invitations, revocation, scoped API keys and OAuth.
Learn more →MySQL and JSON
Server-side MySQL metadata reading, manual synchronization and validated JSON. Other direct connectors are not advertised.
Learn more →Export and saving
schema.json, view.json, PNG/SVG, server history and a local queue during outages.
Learn more →REST and MCP
Shared application services, OpenAPI and generated tool schemas; pagination and revisions.
Learn more →SchemaMap features · SchemaMap documentation
From a task to a saved map
Explore a large existing database
Start with a known legacy database table, follow FK neighbors and save an explainable investigation map.
Map a PHP method’s database tables with Codex
PHP method → a local agent with permitted source access → tables and JOIN evidence → a saved SchemaMap task sheet.
Discuss planned database changes
Compare source and planned task structures on a map and discuss relationships and indexes before applying a migration.
Hand off investigation context
Share goals, sheets, JOIN evidence, notes and plans with a colleague or assistant while preserving access boundaries.
AI and automation
ChatGPT, Claude and Codex use MCP or REST API within granted permissions. An agent investigates code in its permitted environment and submits findings; SchemaMap preserves the map and context. Tool availability and own-proof retention depend on the client.
MCP, REST and AI integrations · Data, access and boundaries
Frequently asked questions
What can I try without registration?
The public shop demo offers table/column search, map viewing, JSON and PNG/SVG. Its changes are local. Open the application with a guest session or Google sign-in to save your own work. Documentation →
Which databases and imports are supported?
Direct server-side MySQL metadata reading and normalized JSON import. PostgreSQL is internal storage; a direct PostgreSQL connector is not advertised. Documentation →
Does SchemaMap modify the source database?
No. The map uses metadata. SQL and migration plans remain descriptive text. Voluntary screenshots and invented examples are stored separately. Documentation →
How do I connect an AI assistant?
Use «API и MCP» in the application: grant minimum OAuth scopes or a separate API key. Source analysis needs a separately permitted agent environment; a web chat does not wake automatically to read replies. Documentation →
Start exploring
Thousands of tables? Start with one.
Database structure is enough to begin. You control data and permissions.