KytheraDB -- Features
Complete feature reference for KytheraDB. Each feature includes its description, the problem it solves, key capabilities, plan availability, and related features.
Table of Contents
- Schema Explorer and Graph Viewer
- Data Explorer (CRUD)
- SQL Console
- Mass Patch
- Bridge / Data Transfer
- Anonymization
- Seed Studio
- Validation and QA
- Performance Monitoring
- Preflight Guard
- RGPD/Compliance
- Import/Export
- SQL Console Features (Terminal v2)
- Configuration Center
- Governance
- Multi-DB Support
- Authentication and Billing
- Smart Synthetic Data
- Custom References
1. Schema Explorer and Graph Viewer
Description: Real-time database introspection that displays every table, column, type, constraint, index, and foreign key in your database. The graph viewer renders table relationships as an interactive directed graph.
Problem it solves: Developers spend significant time reading migration files, ERD tools, or running manual queries to understand database structure. Schema Explorer gives instant, always up-to-date visibility without leaving the platform.
Key capabilities:
- Full schema introspection (tables, columns, types, nullability, defaults, primary keys, foreign keys, indexes, unique constraints)
- Interactive graph visualization powered by Cytoscape.js with dagre layout
- Focus mode: select a table to see only its direct relationships
- Dependency ordering (topological sort for insert/delete order)
- Schema snapshot history with diff comparison
- Health audit: detect missing PKs, orphan tables, naming inconsistencies
- Domain classification: automatic detection of table roles (reference, transactional, audit, join)
- Table detail panel with column metadata, constraints, and annotations
- Migration script generation from schema changes
Plan availability:
| Capability | Starter | Pro | Enterprise |
|---|---|---|---|
| Schema Explorer | Yes | Yes | Yes |
| Graph Viewer | Yes | Yes | Yes |
| Health Audit | No | Yes | Yes |
| Migration Generator | No | No | Yes |
Related features: Data Explorer, Profiling, Validation
2. Data Explorer (CRUD)
Description: A full-featured data browser and editor that lets you view, create, update, and delete records in any table. Includes dynamic form generation, foreign key resolution with dropdown lookups, pagination, sorting, filtering, and 360-degree entity views showing all related records.
Problem it solves: Switching between database clients, writing ad-hoc queries to inspect data, and manually constructing INSERT/UPDATE statements is error-prone and slow. Data Explorer provides a safe, visual interface for everyday data operations.
Key capabilities:
- Paginated data table with sortable columns
- Dynamic form schema generation based on column types
- Foreign key resolution: dropdown selectors populated from referenced tables
- 360-degree entity view: see a record and all records referencing it across other tables
- Record creation with type validation
- Inline record editing
- Record deletion with confirmation
- Read-only mode for Starter plan
Plan availability:
| Capability | Starter | Pro | Enterprise |
|---|---|---|---|
| Read-only access | Yes | Yes | Yes |
| Write access | No | Yes | Yes |
Related features: Schema Explorer, SQL Console, Import/Export
3. SQL Console
Description: An intelligent SQL execution environment with CodeMirror 6 editor, real-time autocomplete, SQL analysis with risk detection, guard rail enforcement, and query preview. Supports SELECT, INSERT, UPDATE, DELETE, and DDL statements with plan-based permissions.
Problem it solves: Running SQL directly against a database is risky without visibility into what the query will affect. The SQL Console analyzes queries before execution, warns about destructive operations, and enforces guard rails to prevent accidental damage.
Key capabilities:
- CodeMirror 6 editor with SQL syntax highlighting
- Schema-aware autocomplete (table names, column names)
- SQL analysis: parse queries with sqlglot, detect statement types, estimate affected rows
- Guard rail enforcement: block or warn on destructive operations based on project configuration
- Query preview: see affected rows before executing
- EXPLAIN plan visualization
- Read-only mode for Starter plan, full write access for Pro and Enterprise
- Multi-tab support with independent editor state
- Detachable terminal: pop out into a separate browser window
- Pill/overlay/terminal layout modes
Plan availability:
| Capability | Starter | Pro | Enterprise |
|---|---|---|---|
| Read-only SQL | Yes | Yes | Yes |
| Write SQL | No | Yes | Yes |
Related features: Preflight Guard, Mass Patch, Performance Monitoring
4. Mass Patch
Description: A controlled workflow for executing large-scale data modifications across your database. Every patch goes through a mandatory pipeline: creation, guard rail validation, dry-run simulation, approval, execution with automatic snapshots, and optional rollback.
Problem it solves: Mass UPDATE or DELETE operations on production data are high-risk. Without dry-run, snapshots, and rollback, a single mistake can cause irreversible data loss. Mass Patch ensures every bulk operation is auditable, reversible, and validated before execution.
Key capabilities:
- Patch creation with SQL (free-form or template-generated)
- SQL analysis: parse and validate SQL, detect statement types and affected tables
- Guard rail validation: check against project-wide policies (max affected rows, blocked patterns, required WHERE clauses)
- Dry-run simulation: execute in a transaction, capture before/after snapshots, then rollback
- Diff viewer: compare before/after state of affected rows
- Automatic snapshot before execution for rollback
- Batch execution with progress tracking via SSE streaming
- Rollback: restore data from pre-execution snapshot
- Approval workflow with status tracking (draft, dry_run_ok, approved, executing, done, rolled_back)
- Export to SQL or Liquibase XML changeset format
- Duplicate patches for iterative refinement
- Template builder for common operations (update column, delete where, nullify)
Plan availability:
| Capability | Starter | Pro | Enterprise |
|---|---|---|---|
| Mass Patch | No | Yes | Yes |
Related features: SQL Console, Preflight Guard, Configuration Center (guard rails)
5. Bridge / Data Transfer
Description: Transfer data between two database connections within the same project. Handles schema differences, FK ordering, conflict resolution, and optional anonymization during transfer.
Problem it solves: Moving data between environments (dev, staging, production) typically requires manual SQL dumps, careful ordering of inserts to respect foreign keys, and handling conflicts when target tables already contain data. Bridge automates the entire process.
Key capabilities:
- Validate source and target connections before transfer
- Preview: compare row counts and detect differences between source and target
- Missing tables detection with automatic structure cloning (CREATE TABLE)
- FK-ordered insert: topological sort ensures parent records are inserted before children
- Conflict strategies: skip existing, overwrite, or merge
- Hash-based change detection: compute row hashes to identify actually changed records
- Transfer history tracking
- Optional anonymization during transfer (Enterprise only)
Plan availability:
| Capability | Starter | Pro | Enterprise |
|---|---|---|---|
| Bridge | No | Yes | Yes |
| Bridge + Anonymize | No | No | Yes |
Related features: Anonymization, Schema Explorer, Import/Export
6. Anonymization
Description: Hash-aware data anonymization that replaces sensitive column values with fake, hashed, masked, nullified, or fixed values. Supports in-place anonymization and copy-to-target mode for creating anonymized database copies.
Problem it solves: Development and QA teams need realistic data but cannot use production data containing PII. Manual anonymization is inconsistent and error-prone. Auralith provides deterministic, hash-aware anonymization that preserves referential integrity.
Key capabilities:
- Configurable anonymization strategies per column:
fake: generate realistic fake data using Faker (names, emails, addresses, etc.)hash: deterministic SHA-256 hash (preserves consistency across tables)mask: partial masking (e.g.,john@email.combecomesj***@e***.com)null: set to NULLfixed: replace with a constant value
- Hash-aware: detect already-hashed columns to avoid double-anonymization
- Preview mode: see anonymized output before applying
- In-place anonymization: modify data directly in the source database
- Copy mode (Enterprise): anonymize while transferring to a different connection
- Configuration persistence: save and reuse anonymization configs per project
- Bulk anonymization across multiple tables
Plan availability:
| Capability | Starter | Pro | Enterprise |
|---|---|---|---|
| In-place anonymization | No | Yes | Yes |
| Copy mode | No | No | Yes |
Related features: Bridge, RGPD/Compliance, Export (anonymized)
7. Seed Studio
Description: Generate test data for your database using three modes: standard seeding (random rows per table), entity-aware seeding (respects FK relationships and insert order), and QA templates (predefined scenarios for repeatable test data).
Problem it solves: Creating realistic test data that respects foreign keys, unique constraints, and business rules is tedious and error-prone. Seed Studio automates this with smart defaults, FK-aware planning, and reusable QA templates.
Key capabilities:
- Standard seeding: generate N rows for a selected table with type-aware random values
- Entity seeding: analyze FK graph, build a seed plan that inserts parent tables first, respect unique constraints
- QA Templates: define reusable seed scenarios with specific data patterns, save as templates, run on demand
- Seed plan viewer: visualize the execution order and row counts per table
- Seed preview: see generated data before inserting
- Seed history: track all seed operations with timestamps, row counts, and template references
- Template library: browse, create, edit, duplicate, and delete QA templates
- Dry-run mode: preview generated rows without inserting
- Scenario runner with labeled runs for easy identification
Plan availability:
| Capability | Starter | Pro | Enterprise |
|---|---|---|---|
| Standard seeding | Yes | Yes | Yes |
| Entity seeding | No | Yes | Yes |
| QA Templates | No | Yes | Yes |
| SQL seeding | No | No | Yes |
Related features: Schema Explorer (FK graph), Validation (verify seeded data)
8. Validation and QA
Description: Define validation rules on your database columns and tables, run scans to detect violations, and track data quality over time through a dashboard with trend visualization.
Problem it solves: Data quality issues (nulls in required fields, invalid formats, broken references, out-of-range values) often go undetected until they cause application bugs. Validation and QA provides continuous monitoring and alerting for data integrity.
Key capabilities:
- Rule types:
not_null,regex,unique,range,enum,custom_sql,foreign_key,length - Rule categories:
column,table,cross_table - Scope filtering: apply rules to specific schemas or table patterns
- Environment tags: activate rules only in certain environments (dev, staging, production)
- Regex library: pre-built patterns for emails, phones, URLs, UUIDs, etc.
- Regex tester: validate regex patterns against sample data
- Scan execution: run all rules against the database, capture violations
- Scan runs history with comparison between runs
- Validation dashboard: violation counts, trends, severity distribution
- Auto-suggest: analyze column data to recommend appropriate rules
Plan availability:
| Capability | Starter | Pro | Enterprise |
|---|---|---|---|
| Validation | No | Yes (5 rules max) | Yes (unlimited) |
Related features: Seed Studio, Data Explorer, RGPD/Compliance
9. Performance Monitoring
Description: Monitor database performance by analyzing slow queries captured by pg_stat_statements, computing criticality scores, running EXPLAIN analysis, and generating index recommendations.
Problem it solves: Slow queries degrade application performance but are often invisible to developers. Performance Monitoring surfaces the most impactful queries, explains why they are slow, and recommends concrete index changes.
Key capabilities:
- Integration with PostgreSQL pg_stat_statements extension
- Slow query identification with configurable thresholds
- Criticality scoring: combine execution time, call frequency, and row impact
- Schema-scoped analysis: filter by PostgreSQL schema
- Table statistics: row counts, dead tuples, sequential vs. index scans
- EXPLAIN analysis: run EXPLAIN ANALYZE on any query
- Index recommendations: suggest missing indexes based on query patterns and table statistics
- Dashboard with top slow queries, table health, and aggregate metrics
- Pre-flight check: verify pg_stat_statements is properly configured
Plan availability:
| Capability | Starter | Pro | Enterprise |
|---|---|---|---|
| Performance Monitoring | No | No | Yes |
Related features: SQL Console, Preflight Guard
10. Preflight Guard
Description: A decision-flow engine that analyzes SQL statements before execution. It evaluates risk level based on statement type, affected tables, WHERE clause presence, row estimates, and environment context, then returns a proceed/warn/block recommendation.
Problem it solves: Executing SQL without understanding its impact leads to accidental data loss or corruption. Preflight Guard acts as an automated code review for SQL, catching dangerous operations before they reach the database.
Key capabilities:
- Multi-database adapter architecture (PostgreSQL, MySQL, SQLite)
- Statement type detection: SELECT, INSERT, UPDATE, DELETE, DROP, TRUNCATE, ALTER, CREATE
- Risk level classification:
safe,low,medium,high,critical - Decision flow:
proceed,warn(with confirmation),block(prevents execution) - Environment-aware analysis: stricter rules for production connections
- Row estimation: EXPLAIN-based row count prediction
- WHERE clause analysis: flag UPDATE/DELETE without WHERE
- Batch analysis: analyze multiple statements in one request
- Analysis history: track all preflight checks per project
- Statistics: aggregate risk distribution over time
Plan availability:
| Capability | Starter | Pro | Enterprise |
|---|---|---|---|
| Preflight Guard | No | No | Yes |
Related features: SQL Console, Mass Patch, Configuration Center
11. RGPD/Compliance
Description: A multi-regulation scanner that analyzes database schemas to detect columns likely containing personally identifiable information (PII) or regulated data. Supports GDPR, HIPAA, PCI-DSS, SOX, and other regulatory frameworks.
Problem it solves: Identifying which columns contain sensitive data across dozens or hundreds of tables is a manual, error-prone process. The compliance scanner automates detection using column name patterns, data type heuristics, and regulation-specific rules.
Key capabilities:
- Multi-regulation support: GDPR, HIPAA, PCI-DSS, SOX, and extensible registry
- Column-level sensitivity detection based on naming patterns and data types
- Risk severity classification per detected column
- Regulation-specific recommendations
- Scan results with table/column/regulation mapping
- Available regulations endpoint for frontend configuration
Plan availability:
| Capability | Starter | Pro | Enterprise |
|---|---|---|---|
| GDPR Scanner | No | Yes (5 regulations max) | Yes (unlimited) |
Related features: Anonymization, Export (anonymized), Configuration Center
12. Import/Export
Description: Import data from CSV and SQL files into your database, and export data in multiple formats including CSV, SQL, Markdown documentation, HTML documentation, and anonymized variants.
Problem it solves: Moving data in and out of databases requires format conversion, schema validation, and careful handling of encoding and delimiters. Import/Export handles all of this with validation steps before execution.
Key capabilities:
- CSV Import: upload CSV files, validate columns against table schema, preview data, execute insert
- SQL Import: upload SQL files, validate syntax, execute statements
- Bulk Import: import multiple CSV files at once, one per table
- CSV Export: export single table or full database to CSV
- SQL Export: export full database as INSERT statements
- Schema SQL Export: export DDL only (CREATE TABLE statements)
- Anonymized Export: export with anonymization applied (Enterprise)
- Markdown Documentation: export schema as Markdown reference
- HTML Documentation: export schema as styled HTML (Enterprise)
Plan availability:
| Capability | Starter | Pro | Enterprise |
|---|---|---|---|
| CSV Export | Yes | Yes | Yes |
| SQL Export | No | Yes | Yes |
| Anonymized Export | No | No | Yes |
| Markdown Doc | No | Yes | Yes |
| HTML Doc | No | No | Yes |
| CSV Import | No | Yes | Yes |
| SQL Import | No | No | Yes |
| Bulk Import | No | No | Yes |
Related features: Anonymization, Bridge, Schema Explorer
13. SQL Console Features (Terminal v2)
Description: The SQL Console terminal is a rich editing environment built on CodeMirror 6 with multiple layout modes and advanced IDE-like features.
Problem it solves: Basic SQL text areas lack the productivity features developers expect from modern code editors. Terminal v2 provides a professional-grade SQL editing experience directly in the browser.
Key capabilities:
- Syntax highlighting: full SQL grammar highlighting with CodeMirror lang-sql
- Autocomplete: schema-aware completions for table names, column names, SQL keywords
- SQL lint: real-time error detection and suggestions
- Multi-tab: multiple independent SQL editors open simultaneously
- Detachable terminal: pop the SQL console into a separate browser window for dual-monitor setups
- Layout modes: pill (compact inline), overlay (floating panel), terminal (full-width bottom panel), detached (separate window)
- One Dark theme: consistent dark mode styling matching the application design
- Query execution: run queries directly from the editor with keyboard shortcuts
- Result display: tabular results with pagination and column sorting
Plan availability: Available on all plans (read-only on Starter, full write on Pro and Enterprise).
Related features: SQL Console, Mass Patch, Preflight Guard
14. Configuration Center
Description: Project-level configuration hub for managing data type mappings, guard rail policies, environment definitions, and security settings.
Problem it solves: Different projects have different requirements for what constitutes a dangerous operation, which environments exist, and how security should be enforced. The Configuration Center centralizes these settings per project.
Key capabilities:
- Data type configuration: define custom type mappings and display preferences
- Guard rails: configure policies that block or warn on specific SQL patterns, row count thresholds, and statement types
- Environment management: define environment tags (dev, staging, production) with risk levels
- Security settings: configure access controls, allowed operations, and audit requirements
- Per-project configuration persistence
- Configuration accessible by admin role only
Plan availability: Available on all plans.
Related features: Mass Patch (guard rails), Preflight Guard (environment awareness), Governance
15. Governance
Description: Enterprise-grade governance dashboard for DSI (IT directors) providing a bird's-eye view of all teams, projects, users, and activity across the platform. Includes team management, project permissions, audit trail, session management, and user blocking.
Problem it solves: In organizations with multiple teams and projects, IT leadership needs visibility into who is doing what, the ability to manage access centrally, and an audit trail for compliance. Governance provides all of this from a super_admin perspective.
Key capabilities:
- DSI Dashboard: aggregate metrics across all projects, teams, users, and activity
- Team management: create, view, and delete teams; manage team members and roles
- Project overview: see all projects with their connections, team assignments, and activity
- Project detail: deep dive into a specific project's audit log, connections, and configuration
- Audit trail: searchable, filterable log of all actions across the platform with CSV export
- Alert system: detect anomalies (high-risk operations, unusual activity patterns)
- Session management: view active sessions, force-logout users
- User management: list all users with their plans, roles, activity; block/unblock users
- Project permissions: grant per-project access (READ, WRITE, EXECUTE, PATCH, ADMIN) to specific users
- Deleted projects tracking: view audit trail for projects that have been removed
Plan availability: Governance is accessible only to super_admin users and is an Enterprise-level feature.
Related features: Teams, Configuration Center, Audit Trail
16. Multi-DB Support
Description: KytheraDB supports connecting to and managing multiple database engines through a unified connector architecture. Each engine has its own introspection adapter, SQL dialect handling, and capability set.
Problem it solves: Teams often work with multiple database technologies. Rather than requiring separate tools for each engine, Auralith provides a consistent interface across all supported databases.
Key capabilities:
- PostgreSQL: full support including pg_stat_statements, EXPLAIN ANALYZE, schemas, advanced types
- MySQL / MariaDB: schema introspection, CRUD, seeding, export, bridge
- SQLite: file-based database support for local development and testing
- SQL Server (Enterprise): Microsoft SQL Server connectivity
- Oracle (Enterprise): Oracle Database connectivity
- Unified connector interface: all engines expose the same introspection and execution API
- SSH tunnel support: connect to databases behind firewalls via SSH
- Connection encryption: credentials stored with Fernet symmetric encryption
- Per-connection environment tagging (dev, staging, production)
Plan availability:
| Engine | Starter | Pro | Enterprise |
|---|---|---|---|
| PostgreSQL | Yes | Yes | Yes |
| MySQL / MariaDB | No | Yes | Yes |
| SQLite | No | Yes | Yes |
| SQL Server | No | No | Yes |
| Oracle | No | No | Yes |
Related features: Schema Explorer, Bridge, Connection Management
17. Authentication and Billing
Description: Complete user authentication with email verification, password reset, Stripe-powered billing with inline payment (no redirect), and subscription management.
Registration:
- Two account types: Particulier (individual) and Entreprise (company with SIRET)
- Email verification via Mailpit (dev) / SMTP (prod)
- Pydantic EmailStr validation on backend
Authentication:
- JWT tokens (24h expiry), Argon2 password hashing
- Email verification required before login
- Password reset via email with expiring tokens
- Rate limiting: 5 requests/minute on login, register, forgot-password, resend-verification
Billing:
- Stripe Elements inline payment (no redirect to Stripe Checkout)
- Plans: Starter (free), Pro (29/mo), Enterprise (99/mo) with annual -20%
- Subscription lifecycle: create, sync, cancel at period end
- Invoice history with PDF download from Stripe
- Automatic plan update after payment confirmation via /billing/sync
Plan availability: All plans
Related features: Governance, Configuration Center
18. Smart Synthetic Data
Description: Domain-aware data generation using French datasets and region-based Faker locales. Replaces generic Faker-only generation with localized, realistic data.
Architecture:
- DatasetProvider: loads JSON datasets (cities, names, streets, companies) by region
- RuleEngine: post-processing coherence (city + postal code matching, phone format)
- Faker fallback: for columns not covered by datasets
- Region selector: France (datasets + Faker fr_FR), or any other country (Faker only with correct locale)
Datasets included (France):
- 100 real cities with postal codes, departments, regions (INSEE)
- 100 common first names, 100 last names
- Street types and 50 common street names
- Company name patterns (prefixes, suffixes, family names)
Supported regions: France, Belgique, Suisse, Canada, USA, UK, Allemagne, Espagne, Italie
Plan availability: All plans
Related features: Seed Studio, Anonymization
19. Custom References
Description: User-defined reference data lists associated to table columns. Feeds into seed generation, anonymization, validation, and data explorer. Replaces the need for hardcoded Faker types when domain-specific values are needed.
Key capabilities:
- Create references tied to table.column with values (manual entry, JSON textarea, file upload)
- Auto-detection of enum candidates from schema analysis (confidence scoring, 70%+ threshold)
- Confirm detected candidates as references
- Import distinct values from the live database
- Upload JSON files with drag-and-drop support (max 500 entries per reference)
- Reference badge in Configuration (Types de donnees) — disables Faker type select when a reference is linked
- Integration with generate-row endpoint — uses references before Faker fallback
- Integration with seed engine — DatasetProvider checks references
- Search/filter across all references
Data flow:
- Detection or manual creation → ReferenceDefinition (confirmed, table.column)
- generate-row, seed, anonymization → check ReferenceDefinition for the column → use values
- Configuration (Types de donnees) → shows badge, disables manual type override
Plan availability: All plans
Related features: Seed Studio, Anonymization, Validation, Configuration Center