Scoutari

AI builder 的決策情報台。

瀏覽

  • 今日情報
  • Skill 榜單
  • AI 週報
  • 關於與編輯方針
  • RSS

主題

  • 模型發布
  • API 與價格
  • MCP 與 Skills
  • AI Coding
  • 中國模型
  • 開源模型
  • Agent
  • Infra / 成本
  • 值得試用
© 2026 Scoutari · scoutari.com摘要由 AI 輔助生成,並始終附上原文連結。
情報收藏週報設定
SScoutariAI Builder Intel · decision desk
Skill RadarSavedPersonalizeSubmit sourceLog inSign up
Sign up
Skill RadarSavedPersonalizeSubmit sourceLog in

Agents

Autonomous agents, tool use, computer use, multi-agent systems, and agent frameworks.

Model releasesAPI & pricingMCP & SkillsAI codingChinese modelsOpen sourceInfra & costMultimodal & imageAI videoVoice AIAI hardwareRobotics & embodiedAI safetyResearch & papersEnterprise AIAI fundingPolicy & regulationWorth trying
  • Mon, July 6Tools

    Agency-Agents-ZH: 266 Plug-and-Play AI Expert Roles with 18 Tools

    GitHub repo agency-agents-zh offers 266 plug-and-play AI expert roles supporting 18 tools across 20 departments, including 50 original agents tailored for the Chinese market.

  • Mon, July 6Open Source

    MaxKB: Open-Source Enterprise Agent Platform Trending on GitHub

    MaxKB is an open-source enterprise agent platform gaining significant traction on GitHub.

  • Mon, July 6Tools

    manaflow-ai/cmux: Ghostty-based macOS Terminal with Vertical Tabs and AI Agent Notifications

    manaflow-ai/cmux is an open-source macOS terminal built on Ghostty, featuring vertical tabs and AI coding agent notifications.

  • Mon, July 6Open Source

    Multica: Open Source Platform to Make Coding Agents Team Members

    Multica is an open-source hosted agent platform that turns coding agents into real team members, supporting task assignment, progress tracking, and skill compounding.

  • Mon, July 6Open Source

    K-Dense-AI/scientific-agent-skills: Turn Any AI Agent into a Scientist

    K-Dense-AI/scientific-agent-skills provides 140 ready-to-use skills and 100+ scientific databases to turn any AI agent into an AI scientist. It already serves 160k scientists.

  • Mon, July 6Open Source

    DeepSeek-Reasonix: Terminal-native AI coding agent with prefix-cache stability

    esengine/DeepSeek-Reasonix is a DeepSeek-native terminal AI coding agent designed for prefix-cache stability.

  • Mon, July 6Tools

    Anthropic-Cybersecurity-Skills Releases 817 AI Agent Cybersecurity Skills

    GitHub trending project mukul975/Anthropic-Cybersecurity-Skills provides 817 structured cybersecurity skills mapped to six security frameworks, usable on 20+ AI platforms.

  • Mon, July 6Tools

    n8n workflow automation platform rises 865 stars on GitHub with native AI capabilities

    n8n is a fair-code workflow automation platform with native AI capabilities, supporting visual building and custom code, self-hosted or cloud, with over 400 integrations, recently gaining 865 stars on GitHub Trending.

  • Mon, July 6Tools

    PM Skills Marketplace Launches with 100+ Agent Skills

    GitHub trending project pm-skills launches PM Skills Marketplace with over 100 agent skills, commands, and plugins.

  • Mon, July 6Open Source

    claude-mem: Persistent context across sessions for all agents

    claude-mem captures all agent actions in a session, compresses them with AI, and injects relevant context into future sessions.

  • Mon, July 6Tools

    coreyhaines31/marketingskills hits 36K+ stars on GitHub, offering CRO, copywriting, SEO tips

    Trending GitHub repo coreyhaines31/marketingskills provides marketing skills for Claude Code and AI Agents, covering CRO, copywriting, SEO, analytics, and growth engineering.

  • Mon, July 6Tools

    Oh My Pi: Terminal AI Coding Agent with Hash-Anchored Edits, LSP, Browser & Subagents

    Oh My Pi is a terminal-based AI coding agent featuring hash-anchored edits, an optimized tool harness, LSP, Python, browser, and subagents.

  • Mon, July 6Tools

    ruvnet/ruflo: Multi-agent meta-harness with Claude Code, multi-model integration

    A meta-harness for deploying multi-agent swarms, coordinating autonomous workflows, and building conversational AI systems.

  • Mon, July 6Tools

    earendil-works/pi: A unified LLM API and Agent loop toolkit

    GitHub trending repo earendil-works/pi offers an AI toolkit with unified LLM API, Agent loop, TUI, and coding agent CLI, gaining 1,423 stars in 7 days (total 67,794).

  • Mon, July 6Agent

    mvanhorn/last30days-skill: Cross-platform AI Agent skill aggregating Reddit, X, YouTube sources for summaries

    A GitHub trending project that researches any topic across multiple platforms and synthesizes evidence-based summaries.

  • Mon, July 6Tools

    CodeGraph: Pre-indexed Code Knowledge Graph, Auto-Syncs Changes, Supports Multiple AI Agents

    CodeGraph is a pre-indexed code knowledge graph that auto-syncs on code changes and supports AI agents like Claude Code, Codex, and Gemini.

  • Mon, July 6Tools

    alirezarezvani/claude-skills releases 337 skills and plugins supporting 11 code agents

    GitHub trending project alirezarezvani/claude-skills offers 337 Claude Code skills, agent skills, and plugins covering engineering, marketing, and more, supporting 11 code agents including Claude Code, Codex, Gemini CLI, and Cursor.

  • Mon, July 6Tools

    Claude Code Hits GitHub Trending, Terminal Agent Coding Tool Gets 136K Stars

    Anthropic's Claude Code, a terminal-based agent coding tool, reaches GitHub trending with 136K stars.

  • Mon, July 6Tools

    addyosmani/agent-skills Hits GitHub Trending, Offers Production-Grade AI Coding Agent Skills

    GitHub trending project addyosmani/agent-skills, with 69,456 stars, provides production-grade engineering skills for AI coding agents.

  • Mon, July 6Tools

    Open Design: Local-First Desktop App Turns Code Agents into Design Engines

    Open Design is an open-source, local-first desktop alternative to Claude Design, enabling code agents to generate designs and export to HTML/PDF/PPTX/MP4.

  • Mon, July 6Open Source

    AgentSkills Specification & Docs Open Source Project Tops GitHub Trending

    The GitHub trending project agentskills/agentskills provides specifications and documentation for Agent Skills, gaining 2262 stars in 7 days.

  • Mon, July 6Tools

    Headroom: Open-Source Tool to Compress Tool Outputs, Logs, Files, and RAG Chunks by 60-95% Tokens Without Quality Loss

    Headroom is an open-source project that compresses tool outputs, logs, files, and RAG chunks before sending them to LLMs, reducing token usage by 60-95% while maintaining answer quality.

  • Mon, July 6Tools

    cc-switch: Cross-platform Desktop AI Code Assistant Hub for CLI Agents like Claude Code, Codex

    GitHub trending project farion1231/cc-switch is a cross-platform desktop all-in-one assistant integrating multiple AI CLI agents.

  • Mon, July 6Open Source

    ECC Launches Agent Performance Optimization Framework for Claude Code, Codex, Cursor, etc.

    GitHub trending project ECC launches as an agent performance optimization system, offering skills, instincts, memory, and security.

  • Mon, July 6Open Source

    OpenMontage: First Open-Source Agent Video Production System, 12 Pipelines, 52 Tools, 500+ Agent Skills

    calesthio/OpenMontage is the world's first open-source, agent-driven video production system with 12 pipelines, 52 tools, and 500+ agent skills, turning AI coding assistants into a full video production studio.

  • Mon, July 6Tools

    JuliusBrussee/caveman: Caveman mode cuts token usage by 65%

    JuliusBrussee/caveman is a Claude Code skill that uses caveman-style speech to reduce token consumption by 65%.

  • Sun, July 5Tools

    LlamaIndex Launches legal-kb Demo: Agent-Based Index v2 Document Knowledge Base

    LlamaIndex releases legal-kb, a reference app enabling AI agents to access Index v2 document knowledge bases via tools like retrieve, find, read, and grep, resembling a file system.

  • Sun, July 5Research

    2026 Guide to Open-Source PDF Structured Extraction Models

    This article explains how to use open-source models to convert enterprise data from PDFs, scans, and slides into structured JSON for LLMs and agents.

  • Sun, July 5Tools

    Anthropic Launches Claude Science Beta: Multi-Agent Workbench for Reproducible Pipelines

    On June 30, 2026, Anthropic released Claude Science Beta, a multi-agent collaboration workbench built on Claude models for reproducible pipelines in genomics, proteomics, and cheminformatics.

  • Sat, July 4Tools

    Claude Code v2.1.200 Fixes Key Background Agent and Permission Mode Issues

    Claude Code v2.1.200 fixes multiple issues with background agents, permission modes, and screen readers, and improves the installation script.

  • Fri, July 3Tools

    ponytail: Let AI Agent Think Like a Lazy Senior Developer, Write Less Code

    GitHub trending project ponytail makes AI agents emulate a lazy senior engineer's mindset, advocating that the best code is code never written.

  • Fri, July 3Agent

    Claude Code v2.1.199 Fixes Sub-Agent & Daemon Stability

    Claude Code v2.1.199 fixes sub-agent error handling, background daemon crashes, and SSL error prompts, improving stability.

  • Fri, July 3Agent

    Alibaba Page Agent: JavaScript GUI Agent for Web Control via Natural Language

    Alibaba's Page Agent reads the DOM directly via client-side JavaScript to execute natural language commands, no screenshots or multimodal models needed.

  • Fri, July 3Tools

    llm-coding-agent 0.1a0: CLI & Python API Coding Agent

    Simon Willison releases llm-coding-agent 0.1a0, a simple coding agent based on the LLM library.

  • Fri, July 3Research

    Using DSPy to Evaluate and Improve Datasette Agent SQL System Prompt

    A study uses DSPy to evaluate and improve Datasette Agent's SQL system prompt.

  • Fri, July 3Tools

    Amazon SageMaker AI Multi-Turn RL Best Practices

    AWS shares best practices for reliable multi-turn RL training in SageMaker AI, covering environment setup, external evaluation, reward design, agent change management, and monitoring.

  • Thu, July 2Tools

    Scritty: Shared Searchable Memory for Every AI Coding Agent

    Scritty provides shared, searchable memory for each AI coding agent.

  • Thu, July 2Agent

    Autoresearch: Feedback Loop Behind Self-Improving Agents

    Introspection co-founder Roland Gavrilescu explains autoresearch, agent 'recipes', self-improvement loops, and the central role of humans in the software factory.

  • Thu, July 2Tools

    Convert Research PDFs to Structured JSON with Lift: Controlled, Schema-Driven Field Evaluation

    This tutorial builds a complete PDF-to-structured-data workflow with Lift, focusing on controlled evaluation.

  • Thu, July 2Tools

    Claude Code v2.1.198: Chrome GA, Agent Notifications, AWS Upstream, Fixes

    Claude Code v2.1.198 makes Chrome's Claude generally available, adds background Agent notifications, /dataviz skill, AWS upstream support, and fixes multiple issues.