thedotmack/claude-mem
Claude-Mem is a plugin for Claude Code that captures, compresses, and preserves coding session context to enable persistent memory and continuity across sessions.
Awesome AI Agents › Continual Learning Agents
baidu-baige/LoongFlow
LoongFlow is an open-source framework from Baidu for building what its authors call expert-grade agents, meaning agents that carry a domain specialist workflow rather than answering single prompts. Two mechanisms define it. The first is the PES paradigm, which structures every cycle into planning, execution and summary phases so that an agent decomposes a long task, acts, and then records what the step produced; the authors present this as the way to keep long-horizon reasoning coherent over many iterations. The second is a multi-structure fusion memory that stores reasoning context generated during runs, so experience gathered in earlier attempts is synthesized and fed back into later ones, producing a run-and-improve loop instead of independent attempts. Three reference agents ship with the framework and demonstrate the pattern in different domains. The general agent handles skill-driven coding work from small applications through to bug hunting. The mathematics agent drives the design and iterative refinement of mathematical algorithms. The machine learning agent constructs a full modelling pipeline autonomously and then evolves it. The evolutionary framing is explicit in the project metadata, which positions it as an alternative to OpenEvolve and lists genetic algorithms, iterative refinement and LLM ensembles among its methods, so candidate solutions are generated, evaluated against a task metric, and selected across generations. The package requires Python 3.12 or later, is published on PyPI, is licensed under Apache 2.0 and is accompanied by an arXiv paper. Documentation and discussions are hosted in the repository, with per-agent directories holding runnable configurations. The intended user is a developer or domain expert who wants to encode professional knowledge into an agent that improves measurably over repeated runs, particularly on algorithm design and machine learning tasks where an objective score exists.
https://github.com/baidu-baige/LoongFlow
Also in Continual Learning Agents
vectorize-io/hindsight
Hindsight is an agent memory system designed to create smarter, learning AI agents, addressing limitations of traditional memory techniques like RAG and knowledge graphs.
langchain-ai/langchainjs
LangChain.js is a TypeScript framework for building context-aware, reasoning applications powered by large language models, offering modular components, chains, agents, and tools for production and deployment.
aden-hive/hive
Hive is an outcome-driven agent development framework designed for evolving agents with AI evaluation, human-in-the-loop, and self-improvement capabilities.
aiming-lab/MetaClaw
Proxy-based layer that turns everyday conversations with a personal agent into learning signals, injecting skills per turn and running scheduled reinforcement learning updates without a local GPU.
kayba-ai/agentic-context-engine
Agentic Context Engine (ACE) is a framework that enables AI agents to autonomously learn from their successes and failures, continuously improving their performance across various tasks without additional training data or fine-tuning.
AgentDock/AgentDock
AgentDock is an open-source framework for building sophisticated AI agents with configurable determinism, enabling reliable and creative AI applications through a node-based architecture and multi-stage workflows.
modelscope/AgentEvolver
Self-evolving agent training framework from ModelScope that combines automatic task generation, experience-guided exploration and attribution-based credit assignment into one reinforcement learning stack.