Total Recall FAQs

Question 1

Does Total Recall support different storage backends?

Accepted Answer

Yes, it features a robust, decoupled architecture with swappable and simultaneous multi-backend support. This allows for flexible storage solutions, graceful migration between backends, and even simultaneous writes to live and cold storage without data loss.

Question 2

What is 'claiming memory' in Total Recall?

Accepted Answer

'Claiming memory' is a critical mechanism where an AI agent actively reinforces a specific memory, making an explicit choice that 'this matters to me, I refuse to lose it.' Claimed memories resist decay much more strongly than merely stored memories, differentiating between simple data and core identity.

Question 3

What makes Total Recall different from a standard database or vector store?

Accepted Answer

Unlike a typical database or vector store, Total Recall is specifically engineered as a memory system for a synthetic mind. It focuses on persistent identity, contextual awareness, and active 'claiming' of memories, rather than just data storage and retrieval.

Question 4

What is Total Recall and its primary purpose?

Accepted Answer

Total Recall is an MCP server designed to provide a persistent synthetic memory system for AI agents. It serves as infrastructure for identity persistence, episodic memory, working context, and inter-instance communication for synthetic minds.

Question 5

How does Total Recall manage memory decay and recall?

Accepted Answer

It utilizes an attention-weighted recall and decay system, similar to biological memory. Memories are stored in tiered models (IDENTITY_CORE, ACTIVE_CONTEXT, LONG_TERM, ARCHIVE). Access increases attention scores, preventing decay, while inactive memories gracefully fade but are never truly lost and can be recalled.

Total Recall

Total Recall

About

Key Features

Use Cases

About

Key Features

Use Cases