total-agent-memory

Overview

total-agent-memory is a sophisticated memory system for AI coding assistants like Claude Code, Codex CLI, and Cursor. It maintains a persistent, local knowledge base that learns how developers work, not just what they said. The system uses temporal knowledge graphs, procedural memory, and AST-based codebase ingestion to provide context across sessions and projects. It features hybrid retrieval combining BM25, dense vectors, graph relations, and other techniques to achieve 96.2% R@5 on LongMemEval benchmarks.

Try asking AI

After installing, here are 3 things you can ask your AI assistant:

• memory_save_fast

  Save memory content quickly without LLM calls in fast mode

  Try:Remember that we chose pgvector over ChromaDB for multi-tenant RLSTry:Save this decision about using JWT-only session tokens
• memory_search_fast

  Search memory quickly with hybrid retrieval in fast mode

  Try:Why did we pick pgvector again?Try:Find information about vector database choices
• memory_explain_search

  Explain how a search query was processed

  Try:Explain how you found that memory about database choicesTry:Show me the search pipeline for this query
• memory_warmup

  Warm up the memory system for faster initial access

  Try:Warm up the memory systemTry:Preload memory caches
• memory_perf_report

  Get performance metrics for the memory system

  Try:Show me memory performance metricsTry:Get latency stats
• memory_rebuild_fts

  Rebuild the full-text search index

  Try:Rebuild the full-text search indexTry:Update search indexes
• memory_rebuild_embeddings

  Rebuild all embedding indexes with current models

  Try:Rebuild embedding indexesTry:Update vector embeddings
• memory_eval_locomo

  Run LoCoMo benchmark on memory system

  Try:Run LoCoMo evaluationTry:Test memory accuracy with LoCoMo
• memory_eval_recall

  Evaluate memory recall performance

  Try:Evaluate memory recallTry:Test how well memory retrieves information
• memory_eval_temporal

  Evaluate temporal reasoning capabilities of memory

  Try:Evaluate temporal memoryTry:Test memory timeline accuracy
• memory_eval_entity_consistency

  Evaluate consistency in entity extraction across sessions

  Try:Test entity consistencyTry:Check how well entities are tracked over time
• memory_eval_contradictions

  Check for contradictions in memory knowledge

  Try:Find contradictions in memoryTry:Check for inconsistent information

Comparable tools

Installation

pip install total-agent-memory

{
  "mcpServers": {
    "total-agent-memory": {
      "command": "python",
      "args": ["-m", "total_agent_memory.server"],
      "env": {}
    }
  }
}

total-agent-memory init
total-agent-memory serve

Compare total-agent-memory with