Tag
Claude Brain is an open-source tool that adds persistent memory to Claude Code with a single file, automatically recording decisions, bugs, and plans, with local search support and no need for a database or cloud service.
This paper introduces dependency-guided rollback repair for memory-augmented agents, a method that builds a typed memory-to-action graph from runtime provenance to selectively undo faulty memory effects while preserving benign state, achieving strong recovery on benchmarks.
Introduces mem0, an open-source AI agent memory layer that gives Claude and other coding agents persistent long-term memory, fixing the issue of starting fresh each chat session.
Mem0 is a persistent memory layer for AI agents, launched on Product Hunt to give agents long-term memory capabilities.
The author shares lessons from embedding an AI agent with persistent memory, tool access, and reflexes into a physical robot, highlighting that quick reflexes must bypass the agent, memory needs consolidation, capability requires restraint, and timing is key to perceived intelligence.
This paper introduces CMP (Cognitive Memory Primitive), a continual-learning architecture that uses sparse relational codes and local learning to reduce catastrophic forgetting, demonstrating better backward transfer than a Transformer with EWC on a byte-level language modeling protocol.
The author argues that big AI labs will offer persistent memory but not true continual learning due to power constraints, keeping continually learning models private.
MemOS is an open-source memory operating system for developers building LLM applications and AI agents, providing persistent, editable memory through a unified API with features like multimodal memory and async ingestion.
A developer built an open-source AI control plane called TormentNexus that integrates over 26,000 MCP servers and supports persistent memory tiers.
This paper introduces shared selective persistent memory for agentic LLM systems, which retains reusable context like task specs and data schemas while discarding session-specific reasoning traces, achieving 96% task completion and significant token cost reductions in enterprise scenarios.
Capn-hook is a CLI tool that gives coding agents persistent memory, saving files that answer questions about a codebase so agents don't re-explore the same mysteries across sessions, reducing token usage by 77% on repeat questions.
An exploration of building specialized 'mental model agents' in Grok using custom skills, tool orchestration, and persistent memory to enable more structured, transparent reasoning on complex problems.
A developer introduces Watch Skill, an open-source MCP server that gives AI agents persistent memory for videos by creating local indexes of transcripts, OCR, and visual observations for efficient retrieval.
A new 130+ page survey on always-on agents, which are systems whose future behavior depends on durable state built up across earlier interactions. It treats state as more than memory, scoring each state item on six axes across a lifecycle, and introduces the Always-On Evaluation Protocol (AOEP-v0) to govern state mutation and recovery.
Explains why production AI agents require a three-tier persistent memory stack—session logs (Zep), personalization (Mem0), and governed context (ContextNest)—to avoid retrieval of stale or conflicting facts.
WorldDirector introduces a framework for controllable video world simulation that decouples semantic motion planning from visual rendering using an LLM to coordinate 3D trajectories and camera movements, enabling persistent object memory and stable appearance even after objects re-enter the scene.
The article discusses the emerging pattern of 'wiki memory' for AI agents, where raw source data is intelligently compressed into a persistent, structured knowledge layer that agents can use efficiently. It compares this to basic RAG and gives examples like DeepWiki and LLM Wiki.
Cognee is an open-source AI memory platform that provides persistent long-term memory for AI agents, supporting knowledge graphs and vector embeddings for cross-session memory and retrieval.
Second Brain is an open-source memory layer that lets Claude, ChatGPT, Cursor, and Codex share persistent, semantic searchable memory, deployed on Cloudflare Workers for user-controlled storage.
A detailed overview of a self-hosted persistent memory system for Claude using MCP, SQL Server, and local embeddings, enabling narrative continuity and identity across sessions.