The memory‑systems feed

314 repositories created on GitHub in the 48 hours to 2026-10-31T23:27:00Z

The exact search: agent memory created:2026-10-29T23:27:00Z..2026-10-31T23:27:00Z

Repositories whose name, description or topics mention both 'agent' and 'memory', created in the window, forks excluded by GitHub's default. Stricter readings are in 'secondary'. Stricter readings are smaller: topic:agent-memory 48h: 35. Response SHA-256 1b1ce616a022

I need a memory system for my agent: Python, MIT licence, runs locally. Which fits, and is anything faster or cheaper?

The answer from stored task objects, with no model call

Memory systems for an agent (Python, MIT licence, runs locally), from the memory-systems feed: scan 20261031T2327Z at 2026-10-31T23:30:18Z, 587 repositories gathered, answered from stored nouns with no model call.
1. semantica-agi/semantica (https://github.com/semantica-agi/semantica): 14358 stars, last commit 2026-10-31, MIT, Python, local+hosted. LongMemEval, LoCoMo, latency, cost: not reported.
2. tigerless-labs/agent-memory (https://github.com/tigerless-labs/agent-memory): 4303 stars, last commit 2026-10-30, MIT, Python, local. LongMemEval, LoCoMo, latency, cost: not reported.
3. OSU-NLP-Group/HippoRAG (https://github.com/OSU-NLP-Group/HippoRAG): 4085 stars, last commit 2026-10-06, MIT, Python, local+hosted. LongMemEval, LoCoMo, latency, cost: not reported.
4. aiming-lab/SimpleMem (https://github.com/aiming-lab/SimpleMem): 3858 stars, last commit 2026-07-24, MIT, Python, local. LongMemEval, LoCoMo, latency, cost: not reported.
5. ClaudioDrews/memory-os (https://github.com/ClaudioDrews/memory-os): 1384 stars, last commit 2026-09-20, MIT, Python, local. LongMemEval, LoCoMo, latency, cost: not reported.
Faster or cheaper? Reported speed or cost: xerj-org/xerj (Rust, Apache-2.0, local) cost "XERJ | 9,982 | $1.58 |"; MemTensor/MemOS (TypeScript, Apache-2.0, local+hosted) cost "72% lower token usage and multi-agent memory sharing"; supermemoryai/supermemory (TypeScript, MIT, local) latency "One call. ~50ms.".
None of the picks above reports latency or cost.
Numbers are the projects' own claims, quoted verbatim from their READMEs or papers and checked character by character; nobody here re-ran them. Ranked by quoted evidence, then stars, then recency.

337 memory systems among 587 repositories gathered

• Long-Term Agent Memory through Complementary Curation and Consolidation 2610.39790v1 · 2026-10-29
• Sticky Collusion: Persistence of LLM Agent Collusion after Incentive Removal and Partner Replacement 2610.39660v1 · 2026-10-29
• MemSteer: Turning User Histories into Supervision for Agent Memory Optimization 2610.39276v1 · 2026-10-29
• Routed Continual Learning 2610.38796v1 · 2026-10-28
• CALM: Conflict-Aware Lifecycle Memory Governance for Long-Horizon Language Agents 2610.38573v1 · 2026-10-28
• Is it my preference? Role Grounding Attack against Agent's Memory via Social Interaction 2610.38209v1 · 2026-10-28
• ETAMem: Decoupling and Aligning Heterogeneous Memory for Long-Term Conversational Agents 2610.37443v1 · 2026-10-27
• EdgePace: Deadline- and Freshness-Aware LLM Scheduling on Unified-Memory Platforms 2610.37307v1 · 2026-10-27
• TrawMem: Reframing Agent Memory as Task-Oriented Role-Adaptive Workspaces 2610.36918v1 · 2026-10-27
• From Voiceprints to User Entities: Online Speaker-Memory Construction for Multi-User Spoken Agents 2610.36713v1 · 2026-10-27
• MarkovSim: an LLM Agent Harness with Bounded Context for Long-Horizon Computational Simulation 2610.35349v1 · 2026-10-26
• Adaptive Memory Index Evolution for Agents 2610.34977v1 · 2026-10-26
• BehaviorMem: Building and Evaluating User Memory from Real-World E-Commerce Behavior Streams 2610.34790v1 · 2026-10-26
• The Right Fact, the Wrong Decision: SciValidityBench for Scientific Agent Memory 2610.34767v1 · 2026-10-26
• Agent Memory as Recommendation: Recommender-System Design for Long-Term Memory of LLM Agents 2610.34673v1 · 2026-10-26
• The Price of Scale in Proactive Memory Recall 2610.34507v2 · 2026-10-25
• FlexA-Mem: Budget-Adaptive Agentic Memory for LLM Agents 2610.33274v1 · 2026-10-23
• Agentic Memory Management Is the Next Frontier of Information Retrieval 2610.32464v1 · 2026-10-23
• AfterDeleteBench: Auditing Post-Deletion Information Persistence in LLM Agents 2610.31366v1 · 2026-10-22
• The Value of Memory in Language Agents 2610.31303v2 · 2026-10-22
• Is Complex Agentic Memory Always Needed? Naive RAG Is Stronger Than We Thought 2610.30755v1 · 2026-10-22
• Governance-State Corruption in Long-Term Memory for LLM Agents: A Cross-Model Stress Test 2610.30698v1 · 2026-10-22
• Why Bother Me Again? Benchmarking Personal Agent Memory for Interaction Preferences 2610.30485v1 · 2026-10-21
• Compaction Laundering: Provenance Destruction in Context-Compacting Coding Agents 2610.29869v1 · 2026-10-21
• MemoryTax: The Bitter Lesson of Agent Memory 2610.29823v1 · 2026-10-21
• Uni-Mem: Learning a Long-Horizon Agentic Memory Manager with Unified Online Policy Optimization 2610.29671v1 · 2026-10-21
• PreCoMem: Predictive Cognitive Memory for Self-Evolving Long-Term Dialogue Agents 2610.29467v1 · 2026-10-21
• Memory-Centric Continual Learning for AI Agents 2610.29357v1 · 2026-10-21
• Memory as Code: Declarative Retrieval over Codified Long-Term Memory 2610.27700v1 · 2026-10-20
• REM: What to Store, When to Recall, What to Forget in Experience Memory for Coding Agents 2610.27377v1 · 2026-10-19
• Magnolia: A Persistent-Memory Agent Framework for Computational Chemistry 2610.27230v1 · 2026-10-19
• Which Agent's Skill Should Change? Group-Relative Credit Assignment for Evolving Multi-Agent Teams 2610.27227v1 · 2026-10-19
• BetaMem: Learning to Remember a Changing World for Embodied Agents 2610.26554v1 · 2026-10-19
• The Geometry of Agent Memory: Forecasting Retrieval-Bound Memory Failures, and Characterizing Its Limits 2610.26433v2 · 2026-10-19
• GLOBE: Geometry-Led Online Behavioral Early Exit via Unit-Sphere Trajectories 2610.25970v1 · 2026-10-19
• Learning Complementary Agent Groups via Task Hypergraphs for Next-POI Recommendation with Missing Modalities 2610.25801v1 · 2026-10-19
• Let Them Sleep: Recursive Self-Pruning Memory for Long-Horizon LLM Agents 2610.25561v1 · 2026-10-18
• MemPrice: Online Memory Allocation via Adaptive Pricing for Long-Horizon Agents 2610.24955v1 · 2026-10-18
• Do Conversational Agents Need Explicit Memory? A Controlled Comparison with Raw-History Retrieval 2610.24671v1 · 2026-10-17
• RL-Guided Temporal Localization for Dual-Channel Retrieval in Long-Horizon Agent Memory 2610.23598v1 · 2026-10-16
• From Memory Hits to Answerable Evidence: Query-Conditioned Memory Compilation for Long-Term Agent Memory 2610.23313v1 · 2026-10-16
• SAFE-Mem: Secure Agentic Memory with Fine-Grained Authority and Governance-Aware Retrieval 2610.22959v2 · 2026-10-16
• From Relevant Memories to Sufficient Evidence: Rethinking Memory Utilization in LLM Agents 2610.22668v1 · 2026-10-15

1. A timer enqueues one scan task within its allowance; a worker with a scoped key claims it and reads only official public APIs.
2. Each repository, paper and quoted claim becomes a task object with its source and the time it was seen, written with compare-and-set.
3. Extracting the benchmark claims and comparing the projects are follow-up tasks, each with its own receipt.