Gel Graph-Relational Database

Gel AI provides a suite of tools to integrate AI capabilities into your applications with minimal effort. It consists of three primary interfaces:

1. ext::ai Extension: A Gel extension that automatically generates embeddings for your data. It is compatible with any provider offering a compatible API, including OpenAI, Mistral AI, and Anthropic.
2. Python Library (gel.ai): Provides direct access to all Gel AI features within Python applications.
3. JavaScript Library (@gel/ai): Provides direct access to all Gel AI features within JavaScript backend applications.

Additionally, the ext::vectorstore extension (powered by pgvector) can be used to replicate vectorstore-style workflows, allowing for the storage and searching of embedding vectors and integration with popular AI frameworks.

Gel provides several ways to interact with database instances, whether hosted locally or on a remote Cloud service. You can interact with Gel through:

• Connections: Establishing links to running Gel instances.
• Projects: A local development workflow that simplifies connection configuration and streamlines tasks for the CLI, language servers, and client libraries.
• Command Line Interface (CLI): A tool for database administration, schema management, and REPL-based data exploration.
• Client Libraries: High-level libraries for JavaScript, Python, Go, and Rust that include integrated connection pooling, management, and transaction retries.

To work effectively with Gel, you need to master three primary components:

1. Schema: Defining your data model using the Schema Definition Language (SDL).
2. EdgeQL: Writing queries using EdgeQL, a next-generation query language designed for clarity and power.
3. Standard Library: Utilizing the built-in type system (scalars, collections, objects) and the library of functions and operators.

For advanced topics, syntax details, and internals, refer to the full Reference documentation.

The gel instance command group provides a suite of tools for managing Gel server instances.

Important Note: Most commands in this group are designed for managed Gel environments and are not intended for managing self-hosted instances. Check the specific documentation for individual commands to confirm compatibility with your setup.

Available management tasks include:

• Lifecycle Management: create, start, stop, restart, destroy.
• Connectivity: link (to connect a remote instance), unlink (to disconnect a remote instance).
• Observability: list (show all instances), status (show status of instances), logs (view instance logs).
• Security: credentials (display credentials), reset-password (reset user passwords).
• Maintenance: upgrade (upgrade installations/instances), revert (revert major upgrades).

A Gel project is a codebase linked to a specific Gel database instance. It is identified by the presence of a gel.toml file in the root directory.

Key Benefits:

• Automatic Connection: CLI commands and client libraries automatically connect to the linked instance without requiring connection flags or credentials (e.g., use gel migrate instead of gel -I my_instance migrate).
• Portability: Teammates can quickly set up matching environments by running initialization commands.
• Security: Separates connection details from code, avoiding hard-coded credentials.

Note: Projects are intended for local development only. In production, always provide instance credentials via environment variables.

EdgeQL is a next-generation, object-relational query language used to interact with Gel databases. It is designed to be more intuitive and concise than SQL by bridging the gap between the relational paradigm and the object-oriented nature of modern programming languages.

Key capabilities include:

• Querying data
• Inserting, updating, and deleting data
• Modifying and introspecting the schema
• Managing transactions
• Performing deep queries through links without explicit JOIN syntax