5arXiv cs.CL (Computation and Language)·
KnowAct-GUIClaw: Self-evolving memory and skill framework for cross-platform GUI agents
Researchers introduce KnowAct-GUIClaw, a Know-Route-Act-Reflect agent framework extending the OpenClaw agent system with cross-platform GUI interaction support across Android, iOS, HarmonyOS, and Windows. The system features an experience-attributable memory system and self-evolving skill library that continuously improves from user interaction history. Using open-source Kimi-2.6 models, the framework achieves 64.1% on the MobileWorld long-horizon benchmark, reportedly outperforming closed-source agentic models including GPT-5.5 and Seed-2.0-Pro. The transferable memory and skill components provide an 8.5% improvement when applied across different base models.
Related guides (3)
Related events (8)
MaskClaw: Edge-Side Privacy Arbitration System for GUI Agents with Behavior-Driven Skill Evolution
MaskClaw is an edge-side privacy arbitration framework for GUI agents that intercepts screenshots before they leave a trusted environment, applying Allow/Mask/Ask decisions based on local visual evidence and user-specific policy memory. The system addresses the gap where static PII detectors miss context-dependent privacy boundaries and cloud-side VLMs may upload raw screens before deciding what to protect. The authors introduce P-GUI-Evo, a new benchmark built from real UI patterns and sanitized labels, and demonstrate that pattern matching, cloud reasoning, and routing alone each exhibit systematic failure modes. The artifact is open-sourced on GitHub.
ManimAgent: Self-evolving multimodal agent with cross-task episodic memory for code generation
ManimAgent is a multimodal agent system that accumulates reflection experience across tasks via a dual-channel Episodic Memory Bank, without weight updates or human-curated seeds. The agent generates Python/Manim animations from scientific paper sections, and a vision-language model scores rendered keyframes to populate positive (success rationales) and negative (failure patterns) memory channels. On a fixed-probe evaluation, Pass@1 improves and reflection rounds decrease as memory grows, outperforming no-memory, RAG, and shuffled-memory baselines. The work addresses a known limitation of single-episode reflection in LLM agents by enabling persistent, self-generated learning across task boundaries.
AutoMem: Automated framework trains LLMs to manage memory as a learnable cognitive skill
AutoMem is a new framework that treats memory management in LLMs as a trainable skill, using two optimization loops: one that iteratively revises memory structure via trajectory review by a strong LLM, and one that distills good memory decisions into direct training signal for the agent model. Evaluated on three long-horizon procedurally generated games (Crafter, MiniHack, NetHack), optimizing memory alone yielded 2x-4x performance improvements, bringing a 32B open-weight model competitive with frontier systems like Claude Opus 4.5 and Gemini 3.1 Pro Thinking. The work draws on cognitive science concepts of metamemory and demonstrates that memory management is an independently learnable, high-leverage capability for long-horizon agentic tasks.
PalmClaw: Open-source native on-device LLM agent framework for mobile phones
PalmClaw is an open-source agent framework that runs LLM agents natively on mobile devices, managing sessions, memory, skills, tools, and the agent loop entirely on-device. Unlike existing mobile agents that rely on GUI actions (tapping, swiping), PalmClaw exposes device capabilities as structured tools with explicit arguments and execution boundaries, enabling more direct and controlled access to device features. Experiments report an 11.5% relative improvement in task success rate and a 94.9% reduction in completion time over the strongest baseline.
SkillKit: Portable Skills Layer for AI Coding Agents
SkillKit is an open-source TypeScript project that provides a portable skills abstraction for AI coding agents, enabling installation, translation, and sharing of skills across tools like Claude Code, Cursor, Codex, GitHub Copilot, and 40+ others. The project has accumulated 1,112 stars with 32 added today, indicating moderate community traction. It targets the interoperability gap between the growing ecosystem of AI coding assistants.
UniClawBench: Capability-driven benchmark for proactive agents on real-world tasks
Researchers from HKU MMLab introduce UniClawBench, a benchmark for evaluating proactive AI agents across 400 bilingual real-world tasks organized around five foundational capabilities: Skill Usage, Exploration, Long-Context Reasoning, Multimodal Understanding, and Cross-Platform Coordination. Unlike prior benchmarks, it evaluates agents in live Docker containers with step-by-step checkpoints and a closed-loop multi-turn evaluation strategy using executor, supervisor, and user agents. The benchmark disentangles base model capabilities from agent framework design choices by testing state-of-the-art models across multiple frameworks. Code and benchmark are publicly released.
Hermes Agent Surpasses OpenClaw in Daily Token Usage, Highlighting Self-Improving Agentic Capabilities
Hermes Agent, an open-source personal agent from Nous Research launched in February 2026, has overtaken OpenClaw on OpenRouter's daily token consumption leaderboard. It distinguishes itself through automatic skill creation (converting successful task completions into reusable SKILL.md instruction files), a two-tier memory architecture with intelligent deduplication and merging, and a Curator background process that manages skill lifecycle. The agent supports local or cloud deployment, integrates with ~20 messaging services, and works with a wide variety of LLMs, positioning it as a model-agnostic alternative in the emerging personal agent category.
MobileGym: Verifiable Parallel Simulation Platform for Mobile GUI Agent Training
MobileGym is a browser-hosted simulation environment for mobile GUI agent research that enables deterministic outcome verification via structured JSON state and scalable online RL through hundreds of parallel instances (~400 MB/instance, ~3s cold start). The accompanying MobileGym-Bench provides 416 parameterized task templates across 28 apps with deterministic judges. A sim-to-real case study using GRPO on Qwen3-VL-4B-Instruct achieves +12.8 percentage points on the 256-task test set, with real-device execution retaining 95.1% of simulation-side training gains.