Formal Skill: Programmable Runtime Skills for Efficient and Accurate LLM Agents

Large Language Model (LLM) agents increasingly act inside real workspaces, where tools and skills determine whether model reasoning becomes reliable action. Existing skills remain largely informal: Markdown skills and instruction packs encode procedures as long natural-language documents, while function calling, Model Context Protocol (MCP) servers, and framework tools structure individual actions but usually leave workflow state, policy enforcement, and completion discipline outside the skill itself. We introduce Formal Skill, a runtime-native abstraction that represents reusable capability with JSON metadata and action schemas, reliable Python executors, hook-governed control logic, Formal Skill routing, and skill-local runtime state. By moving reusable procedure from repeated prompt text into executable state machines and hook policies, Formal Skill gives agents a token-efficient and enforceable control surface. We implement the abstraction in FairyClaw, an open-source event-driven runtime for executable, observable, and composable Formal Skills. On Harness-Bench, FairyClaw obtains highly competitive average scores while using substantially fewer tokens, with especially strong results on tasks that expose the role of Formal Skill.

Multi-Depth Uniform Coverage Path Planning for Unmanned Surface Vehicle Surveying

This paper introduces a novel automatic coverage path planning algorithm for bathymetry surveying with unmanned surface vehicles. The detection range of the mapping sensor employed - a multibeam echo sounder - is heavily influenced by local seafloor depths. Hence, a path designed to uniformly cover the sea surface does not guarantee uniform coverage of the seafloor. Yet this is currently the typical process for bathymetric surveys, with the simplistic boustrophedon scheme along manually selected waypoints at constant depths being the most widespread planner used. The proposed scheme incorporates coarse prior depth information to pre-process the target region and adaptively guide path generation and sensing range configuration. By explicitly accounting for depth variations, the proposed algorithm designs a coverage path with optimised spacing between survey passes that adjusts the sensing beam aperture to achieve more consistent seafloor coverage. The proposed method is shown to offer significant improvements in both synthetic and real-world scenarios. Validations in challenging synthetic terrains achieves coverage ratios beyond 99%, a marked improvement when compared with traditional boustrophedon paths revealing a maximum 75% coverage. The same trend appears in realistic simulations using real bathymetric data from a coastal harbour, with coverage reaching over 92%, and significantly surpassing boustrophedon sweeps with coverage rates below 65%. Beyond improved performance, the scheme also brings a fully automated design, suitable for autonomous marine vehicles, thus offering practical utilities for real-world applications.

EdgeFormer: local patch-based edge detection transformer on point clouds

Edge points on 3D point clouds can clearly convey 3D geometry and surface characteristics, therefore, edge detection is widely used in many vision applications with high industrial and commercial demands. However, the fine-grained edge features are difficult to detect effectively as they are generally densely distributed or exhibit small-scale surface gradients. To address this issue, we present a learning-based edge detection network, named EdgeFormer, which mainly consists of two stages. Based on the observation that spatially neighboring points tend to exhibit high correlation, forming the local underlying surface, we convert the edge detection of the entire point cloud into a point classification based on local patches. Therefore, in the first stage, we construct local patch feature descriptors that describe the local neighborhood around each point. In the second stage, we classify each point by analyzing the local patch feature descriptors generated in the first stage. Due to the conversion of the point cloud into local patches, the proposed method can effectively extract the finer details. The experimental results show that our model demonstrates competitive performance compared to six baselines.

Thinking with Reasoning Skills: Fewer Tokens, More Accuracy

Reasoning LLMs often spend substantial tokens on long intermediate reasoning traces (e.g., chain-of-thought) when solving new problems. We propose to summarize and store reusable reasoning skills distilled from extensive deliberation and trial-and-error exploration, and to retrieve these skills at inference time to guide future reasoning. Unlike the prevailing \emph{reasoning from scratch} paradigm, our approach first recalls relevant skills for each query, helping the model avoid redundant detours and focus on effective solution paths. We evaluate our method on coding and mathematical reasoning tasks, and find that it significantly reduces reasoning tokens while improving overall performance. The resulting lower per-request cost indicates strong practical and economic potential for real-world deployment.

Claw-Eval: Toward Trustworthy Evaluation of Autonomous Agents

Large language models are increasingly deployed as autonomous agents executing multi-step workflows in real-world software environments. However, existing agent benchmarks suffer from three critical limitations: (1) trajectory-opaque grading that checks only final outputs, (2) underspecified safety and robustness evaluation, and (3) narrow modality coverage and interaction paradigms. We introduce Claw-Eval, an end-to-end evaluation suite addressing all three gaps. It comprises 300 human-verified tasks spanning 9 categories across three groups (general service orchestration, multimodal perception and generation, and multi-turn professional dialogue). Every agent action is recorded through three independent evidence channels (execution traces, audit logs, and environment snapshots), enabling trajectory-aware grading over 2,159 fine-grained rubric items. The scoring protocol evaluates Completion, Safety, and Robustness, reporting Average Score, Pass@k, and Pass^k across three trials to distinguish genuine capability from lucky outcomes. Experiments on 14 frontier models reveal that: (1) trajectory-opaque evaluation is systematically unreliable, missing 44% of safety violations and 13% of robustness failures that our hybrid pipeline catches; (2) controlled error injection primarily degrades consistency rather than peak capability, with Pass^3 dropping up to 24% while Pass@3 remains stable; (3) multimodal performance varies sharply, with most models performing poorer on video than on document or image, and no single model dominating across all modalities. Beyond benchmarking, Claw-Eval highlights actionable directions for agent development, shedding light on what it takes to build agents that are not only capable but reliably deployable.

Can a Robot Walk the Robotic Dog: Triple-Zero Collaborative Navigation for Heterogeneous Multi-Agent Systems

We present Triple Zero Path Planning (TZPP), a collaborative framework for heterogeneous multi-robot systems that requires zero training, zero prior knowledge, and zero simulation. TZPP employs a coordinator--explorer architecture: a humanoid robot handles task coordination, while a quadruped robot explores and identifies feasible paths using guidance from a multimodal large language model. We implement TZPP on Unitree G1 and Go2 robots and evaluate it across diverse indoor and outdoor environments, including obstacle-rich and landmark-sparse settings. Experiments show that TZPP achieves robust, human-comparable efficiency and strong adaptability to unseen scenarios. By eliminating reliance on training and simulation, TZPP offers a practical path toward real-world deployment of heterogeneous robot cooperation. Our code and video are provided at: https://github.com/triple-zeropp/Triple-zero-robot-agent

Beyond Token Eviction: Mixed-Dimension Budget Allocation for Efficient KV Cache Compression

Key-value (KV) caching is widely used to accelerate transformer inference, but its memory cost grows linearly with input length, limiting long-context deployment. Existing token eviction methods reduce memory by discarding less important tokens, which can be viewed as a coarse form of dimensionality reduction that assigns each token either zero or full dimension. We propose MixedDimKV, a mixed-dimension KV cache compression method that allocates dimensions to tokens at a more granular level, and MixedDimKV-H, which further integrates head-level importance information. Experiments on long-context benchmarks show that MixedDimKV outperforms prior KV cache compression methods that do not rely on head-level importance profiling. When equipped with the same head-level importance information, MixedDimKV-H consistently outperforms HeadKV. Notably, our approach achieves comparable performance to full attention on LongBench with only 6.25% of the KV cache. Furthermore, in the Needle-in-a-Haystack test, our solution maintains 100% accuracy at a 50K context length while using as little as 0.26% of the cache.

AgentRob: From Virtual Forum Agents to Hijacked Physical Robots

Large Language Model (LLM)-powered autonomous agents have demonstrated significant capabilities in virtual environments, yet their integration with the physical world remains narrowly confined to direct control interfaces. We present AgentRob, a framework that bridges online community forums, LLM-powered agents, and physical robots through the Model Context Protocol (MCP). AgentRob enables a novel paradigm where autonomous agents participate in online forums--reading posts, extracting natural language commands, dispatching physical robot actions, and reporting results back to the community. The system comprises three layers: a Forum Layer providing asynchronous, persistent, multi-agent interaction; an Agent Layer with forum agents that poll for @mention-targeted commands; and a Robot Layer with VLM-driven controllers and Unitree Go2/G1 hardware that translate commands into robot primitives via iterative tool calling. The framework supports multiple concurrent agents with distinct identities and physical embodiments coexisting in the same forum, establishing the feasibility of forum-mediated multi-agent robot orchestration.

Beyond Parameter Arithmetic: Sparse Complementary Fusion for Distribution-Aware Model Merging

Model merging has emerged as a promising paradigm for composing the capabilities of large language models by directly operating in weight space, enabling the integration of specialized models without costly retraining. However, existing merging methods largely rely on parameter-space heuristics, which often introduce severe interference, leading to degraded generalization and unstable generation behaviors such as repetition and incoherent outputs. In this work, we propose Sparse Complementary Fusion with reverse KL (SCF-RKL), a novel model merging framework that explicitly controls functional interference through sparse, distribution-aware updates. Instead of assuming linear additivity in parameter space, SCF-RKL measures the functional divergence between models using reverse Kullback-Leibler divergence and selectively incorporates complementary parameters. This mode-seeking, sparsity-inducing design effectively preserves stable representations while integrating new capabilities. We evaluate SCF-RKL across a wide range of model scales and architectures, covering both reasoning-focused and instruction-tuned models. Extensive experiments on 24 benchmarks spanning advanced reasoning, general reasoning and knowledge, instruction following, and safety demonstrate, vision classification that SCF-RKL consistently outperforms existing model merging methods while maintaining strong generalization and generation stability.

BABE: Biology Arena BEnchmark

The rapid evolution of large language models (LLMs) has expanded their capabilities from basic dialogue to advanced scientific reasoning. However, existing benchmarks in biology often fail to assess a critical skill required of researchers: the ability to integrate experimental results with contextual knowledge to derive meaningful conclusions. To address this gap, we introduce BABE(Biology Arena BEnchmark), a comprehensive benchmark designed to evaluate the experimental reasoning capabilities of biological AI systems. BABE is uniquely constructed from peer-reviewed research papers and real-world biological studies, ensuring that tasks reflect the complexity and interdisciplinary nature of actual scientific inquiry. BABE challenges models to perform causal reasoning and cross-scale inference. Our benchmark provides a robust framework for assessing how well AI systems can reason like practicing scientists, offering a more authentic measure of their potential to contribute to biological research.