2026年8月7日 技术热点总结

项目简介：RealReplicaBench: Benchmarking Long-Horizon Agents in High-Fidelity, Stateful, and Reproducible Replicas of Real Online Services

技术栈：HTML

核心介绍：Developed and maintained by the Accio team at Alibaba International. Overview · Live leaderboard · Mock showcase · Quick start · Reproducibility · Contact We run models on re

项目数据：⭐ 1,036 Stars，🍴 74 Forks

核心介绍：将一张照片转化为“原始摄影区域 + 抽象记忆面板 + 诗意英文标题”的竖向编辑作品的 Codex Skill。它保留照片的真实内容，并仅从照片本身提炼空间关系、构图节奏和色彩关系；它不是滤镜、照片重画或风格迁移。 The skill includes the complete prompt in both Chinese and English.

项目数据：⭐ 951 Stars，🍴 52 Forks

项目简介：Locate a nearby Bluetooth device by signal strength, from the macOS command line — for when Find My isn’t available

技术栈：Swift

核心介绍：Locate a nearby Bluetooth device by signal strength, from the command line. Built for the case where Find My is unavailable — for example when a device is enrolled in MDM that disables it — but the device is still within Bluetooth range and you just need to know which corner of the room it is in.

项目数据：⭐ 828 Stars，🍴 134 Forks

项目简介：Find agentic growth hacking skills for Claude, ChatGPT, Manus | by enso.bot

技术栈：Shell

核心介绍：A curated, categorized directory of open-source AI agent skills for growth hacking, marketing execution, and revenue operations. This collection focuses on Agentic Growth Hacking—leveraging AI agents (like Claude Code, Cursor, and OpenClaw) to scale go-to-market workflows, discover attention loopholes, and execute at machine speed. Broad, multi-discipline libraries that cover the entire marketing stack.

关键特性：[Marketing Skills for AI Agents](https://github.com/coreyhaines3。
…

Powered by advances in LLMs and autonomous agents, deep research has become one of the most widely adopted agentic products. However, most deep research systems write general-purpose reports, which are inadequate for financial deep research. Financial research demands specialized knowledge to analyze historical patterns and forecast upcoming events. Automating financial deep research therefore requires both a layered harness to drive the research agent and a verifiable, point-in-time benchmark that prevents leakage of future information. We present FinanceHarness, a harness that runs financ…

Red-team evaluations of AI models support some claims and not others, and the boundary between the two is calculable rather than merely a matter of judgment. We define the evidential ceiling of an evaluation as the largest factor by which one result can move belief under a fixed testing budget, derive it in closed form for the benchmark null result, and use it to locate that boundary exactly. We find that above a calculable harm rate, a benchmark of modest size certifies a category to a stated evidentiary standard, and a clean sheet is then the stronger of the two possible observations, out…

Multimodal large language models (MLLMs) make grounded predictions in real-world scenes by combining visual and textual cues, yet existing benchmarks rarely reveal how they arbitrate between these evidence sources when they conflict. We introduce SIGNPOST-Bench, a controlled counterfactual benchmark for evaluating text-vision conflict resolution. Each source image is transformed into a counterfactual quintuplet of Original, Blank, Similar, Random, and Adversarial variants. Synthetic, localized scene-text interventions are designed to preserve non-textual content, enabling paired measurement…

Model checkpoints are growing in both number and size, which makes archival, transfer, and deployment increasingly costly. General-purpose compressors can reduce storage requirements but ignore tensor structure, whereas existing tensor-specific compressors rely on fixed and format-specific pipelines. We present Brevis, which formulates lossless tensor compression as program synthesis. We design a typed domain-specific language (DSL) that captures recurring tensor structures, such as repeated regions and floating-point fields, through a set of reversible operators. Given a tensor, Brevis syn…

A framework that persists execution state so a run can be interrupted, survive a crash, and continue must decide what a resume means for effects that already fired. Five widely deployed agent workflow frameworks answer differently, none exposes a machine-checkable contract, and behavior violates even the fragments they state. The RESUME CONTRACT states six properties over the persistence API (prefix continuation, effect exactly-once, fork determinism, checkpoint validity, consume-once, recovery determinism), plus fork-intent and liveness obligations. A TLA+ model checks a reference semantic…

Flow-matching vision-language-action (VLA) models such as pi0 generate robot actions by integrating a learned denoising velocity field, and have been reported to resist adversarial perturbations that readily fool autoregressive VLAs. We show that this robustness is largely illusory: it stems from prior attacks ignoring the multi-step denoising ODE. We introduce DRIFT (Denoising Redirection via Input perturbation of the Flow-matching Trajectory), a test-time universal adversarial patch placed on the robot's gripper that attacks the denoising velocity field of an off-the-shelf policy. Our cen…