Go from Message class to ChatCompletion message object

Send a completion request and parse JSON response. Args: messages: List of message dicts with role and content. tempe

Async factory: create a MetaAgent and run async initialization.

Delete a source from the archival memory.

Get the current request context. Returns: The current RequestContext if available, None otherwise.

Retrieve the metadata and content of a node. Args: id: Node identifier. include_embedding: with/without embed

Permanently removes the record from the database (async).

Load trap event data from JSON file.

Send a message to Kafka topic with user_id as partition key. Ensures all messages for the same user go to the same partition,

resolveHubClient 的 viewer 版本：hub 模式下使用 bootstrap admin 身份

urljoin is stupid and wants a trailing / at the end of the endpoint address, or it will chop the suffix off

Compile a SQLAlchemy query to a SQL string for inspection.

Ensure a template block exists for (user, scope, label). Idempotent. If a block with the same (user_id, scope, label) already exists

Build result statistics.

Generic helper to call /product/add and print the payload + response. Args: name: Logical name of this example, printed in logs.

Check whether all filter_tags values can be handled by Redis TAG queries. Returns False if any value is: - A dict with $ operators (Re

Resets the history for a new episode.

Clear all memories. Note: This requires manual database file removal and app restart. Returns: Dict with warnin

Create LLM client.

Delete all textual memories from a MemCube for user. Args: mem_cube_id (str): The identifier of the MemCube to delete th

Get the admin user ID associated with a client.

Get the total number of items in the episodic memory for the user. Args: user: User who owns the memories to count

Get current trace context. Returns: Dictionary with current trace IDs and metadata

Permanently removes the record from the database.

Parse temporal expressions from natural language text. This function identifies temporal expressions like "today", "yesterday", "last week"

Set value (JSON) in Redis.

Create a block with explicit scope in filter_tags; optional extra_filter_tags merged at creation.

Validate required parameters - if passed, they must not be empty

Establish connection to Redis and set up the queue.

Create a new memory cube for a user. Memory cubes are containers that store different types of memories (textual, activation

Create OpenSearch index with proper mappings if it doesn't exist.

Convert a ChatCompletion message object into a Message object (synced to DB)

Extract entities from text using an LLM.

Generate a test ID matching Mirix ID pattern (prefix-[8 hex chars]).

Get user by ID. Args: user_id (str): The user ID. Returns: User: The user object or None if not found.

Get the current UTC time

Check if agent matches any of the given types (by enum or name fallback). Args: *agent_types: One or more AgentType enum value

List all episodic events with various search methods and optional temporal filtering. Args: agent_state: The agent state

Test Redis connection.

Rerank memory items by relevance to task. Args: query (str): Original task. query_embedding(list[float]): em