Generate visualization configuration for query results. Args: request: FastAPI request object id: ChatSessio

Get a value from the cache, respecting TTL

* Get schema metadata for a datasource * @param id - Datasource ID * @returns Database schema information

Get a chat session by ID using the repository. Args: id: Chat session ID Returns: The chat session

Get all datasources with filtering and pagination. Args: skip: Number of records to skip limit: Maximum numb

Get all projects with pagination

Get global async Qdrant manager instance

Get all projects for a datasource

Get a datasource by name. Args: name: Name of the datasource Returns: Datasource if found, None oth

Get all projects for an owner

Get all models for a project

根据 Refresh Token 获取会话

Get a cache instance of the specified type. This factory function returns a singleton instance of the requested cache type. Args:

Get a test value from the cache. Args: requiest: FastAPI request object id: Cache identifier tes

Generate a unique cache key for the request based on path and relevant query params. Excludes certain transient query parameters lik

Example: TODO: Add example This method is used to generate a prompt for the LLM to generate chart spec Args: initial_prompt (st

Get a database connection from the pool with intelligent reuse. Args: db_type: Type of database connection_i

Get context summary for session Args: session_id: Session ID max_messages: Max messages to include

Get context summary for LLM

Generate prompt for SQL correction. Args: question: The original question. wrong_sql: The failed SQL query. error_me

从 Access Token 获取当前用户

Get a datasource by ID with optimized query. Args: db: Database session datasource_id: ID of the datasource

Get all datasources with efficient filtering and pagination. Uses repository pattern for improved database access. Args:

Get synchronous PostgreSQL database instance. Returns: PostgresDB instance

Get a SQLAlchemy database session for sync operations. Yields: SQLAlchemy Session

Get current rate limiter metrics

Get global metrics collector instance. Returns: Singleton MetricsCollector

Get model by ID

Get a database connection from the pool with proper resource management. This context manager handles getting and returning connections to t

Get project by ID

Get global Qdrant manager instance

Get or create global session manager

Example: ```python get_sql_prompt( question="What are the top 10 customers by sales?", few_shots=[{"question": "What are the

Get statistics about the connection pool.

Get list of tables in the database with improved error handling. Returns: List of table names Raises:

Check Qdrant health

Initialize database connections and verify functionality. This function should be called during application startup. Returns: b

Initialize the default PostgreSQL datasource from Docker configuration. Creates a datasource pointing to the Docker PostgreSQL instance

检查 Token 是否过期

List models for a project

List projects with optional filters

Configure logger with diagnostics enabled for debugging.

用户登录

登出（删除 Refresh Token）

Main entry point.

Mask sensitive information in connection_info for API responses. Args: connection_info: Original connection info

刷新 Access Token

Release a connection back to the pool. Args: conn_id: Connection ID to release

Reset the database by dropping and recreating it.

Retrieve similar historical questions Args: question: Current user question project_id: Project ID

Run migrations in 'offline' mode. This configures the context with just a URL and not an Engine, though an Engine is acceptable here

Run migrations in 'online' mode. In this scenario we need to create an Engine and associate a connection with the context.

Generate reasoning plan with streaming output Args: question: User's natural language question mdl_context:

Set a value in Redis with optional TTL

Set a test value in the cache. Args: id: Cache identifier value: Value to store Returns:

* Test connection to a datasource * @param data - Connection details to test * @returns Connection status

Test that IntentAgent correctly parses various formats of 'options' returned by the LLM (dspy module)

Test that concrete business questions are NOT marked as ambiguous.

Test that truly vague questions ARE marked as ambiguous.

Context manager for tracing a single LLM call. Args: name: Name of the LLM call (e.g., "QueryReasoningAgent")

Trace a streaming LLM call. Args: name: Name of the LLM call model: Model name prompt: Input prom

Context manager for transaction handling. Usage: async with adapter.transaction(): await adapter.execute

Update configuration settings.

* Update an existing datasource * @param id - Datasource ID * @param data - Datasource update data * @returns Updated datasource

Update an existing datasource with optimized connection validation. Args: db: Database session datasource_id

Update an MDL model

Update an MDL project

更新用户（仅管理员或本人）

Insert or update points in collection

Validates the given Python object according to the associated schema type. Returns a `Success[T]` object containing the object if validation

验证密码