Starred repositories
τ-Bench: A Benchmark for Tool-Agent-User Interaction in Real-World Domains
A high-throughput and memory-efficient inference and serving engine for LLMs
OpenClaw-RL: Train any agent simply by talking
A browser-based desktop where AI Agent operates every app through natural language.
SkillsBench evaluates how well skills work and how effective agents are at using them.
The RL Bridge for LLM-based Agent Applications. Made Simple & Flexible.
Agent Skills to help developers using AI agents with Supabase
A LLM-based Agent that predict its tasks proactively.
The agent-native LLM router for autonomous agents. 55+ models (8 free), <1ms local routing, USDC payments on Base & Solana via x402.
Research of DeepSeek Engram Architecture based on Qwen-3 and Stable Diffusion series.
Persistent Context Across Sessions for Every Agent – Captures everything your agent does during sessions, compresses it with AI, and injects relevant context back into future sessions. Works with C…
Your own personal AI assistant. Any OS. Any Platform. The lobster way. 🦞
Kimi K2 is the large language model series developed by Moonshot AI team
Dr. Zero Self-Evolving Search Agents without Training Data
SimpleMem: Efficient Lifelong Memory for LLM Agents — Text & Multimodal
PhD Thesis work -- computational model of learning and memory in decision making in reinforcement learning tasks
The code for NeurIPS 2025 paper "A-Mem: Agentic Memory for LLM Agents"
Conditional Memory via Scalable Lookup: A New Axis of Sparsity for Large Language Models
slime is an LLM post-training framework for RL Scaling.