{
  "schemaVersion": "1.0",
  "item": {
    "slug": "embodied-ai-news",
    "name": "Embodied Ai News",
    "source": "tencent",
    "type": "skill",
    "category": "开发工具",
    "sourceUrl": "https://clawhub.ai/HeXavi8/embodied-ai-news",
    "canonicalUrl": "https://clawhub.ai/HeXavi8/embodied-ai-news",
    "targetPlatform": "OpenClaw"
  },
  "install": {
    "downloadMode": "redirect",
    "downloadUrl": "/downloads/embodied-ai-news",
    "sourceDownloadUrl": "https://wry-manatee-359.convex.site/api/v1/download?slug=embodied-ai-news",
    "sourcePlatform": "tencent",
    "targetPlatform": "OpenClaw",
    "installMethod": "Manual import",
    "extraction": "Extract archive",
    "prerequisites": [
      "OpenClaw"
    ],
    "packageFormat": "ZIP package",
    "includedAssets": [
      "SKILL.md",
      "references/news_sources.md",
      "references/output_templates.md",
      "references/search_queries.md",
      "references/taxonomy.md",
      "references/workflow.md"
    ],
    "primaryDoc": "SKILL.md",
    "quickSetup": [
      "Download the package from Yavira.",
      "Extract the archive and review SKILL.md first.",
      "Import or place the package into your OpenClaw setup."
    ],
    "agentAssist": {
      "summary": "Hand the extracted package to your coding agent with a concrete install brief instead of figuring it out manually.",
      "steps": [
        "Download the package from Yavira.",
        "Extract it into a folder your agent can access.",
        "Paste one of the prompts below and point your agent at the extracted folder."
      ],
      "prompts": [
        {
          "label": "New install",
          "body": "I downloaded a skill package from Yavira. Read SKILL.md from the extracted folder and install it by following the included instructions. Tell me what you changed and call out any manual steps you could not complete."
        },
        {
          "label": "Upgrade existing",
          "body": "I downloaded an updated skill package from Yavira. Read SKILL.md from the extracted folder, compare it with my current installation, and upgrade it while preserving any custom configuration unless the package docs explicitly say otherwise. Summarize what changed and any follow-up checks I should run."
        }
      ]
    },
    "sourceHealth": {
      "source": "tencent",
      "slug": "embodied-ai-news",
      "status": "healthy",
      "reason": "direct_download_ok",
      "recommendedAction": "download",
      "checkedAt": "2026-05-02T22:25:01.230Z",
      "expiresAt": "2026-05-09T22:25:01.230Z",
      "httpStatus": 200,
      "finalUrl": "https://wry-manatee-359.convex.site/api/v1/download?slug=embodied-ai-news",
      "contentType": "application/zip",
      "probeMethod": "head",
      "details": {
        "probeUrl": "https://wry-manatee-359.convex.site/api/v1/download?slug=embodied-ai-news",
        "contentDisposition": "attachment; filename=\"embodied-ai-news-1.0.4.zip\"",
        "redirectLocation": null,
        "bodySnippet": null,
        "slug": "embodied-ai-news"
      },
      "scope": "item",
      "summary": "Item download looks usable.",
      "detail": "Yavira can redirect you to the upstream package for this item.",
      "primaryActionLabel": "Download for OpenClaw",
      "primaryActionHref": "/downloads/embodied-ai-news"
    },
    "validation": {
      "installChecklist": [
        "Use the Yavira download entry.",
        "Review SKILL.md after the package is downloaded.",
        "Confirm the extracted package contains the expected setup assets."
      ],
      "postInstallChecks": [
        "Confirm the extracted package includes the expected docs or setup files.",
        "Validate the skill or prompts are available in your target agent workspace.",
        "Capture any manual follow-up steps the agent could not complete."
      ]
    },
    "downloadPageUrl": "https://openagent3.xyz/downloads/embodied-ai-news",
    "agentPageUrl": "https://openagent3.xyz/skills/embodied-ai-news/agent",
    "manifestUrl": "https://openagent3.xyz/skills/embodied-ai-news/agent.json",
    "briefUrl": "https://openagent3.xyz/skills/embodied-ai-news/agent.md"
  },
  "agentAssist": {
    "summary": "Hand the extracted package to your coding agent with a concrete install brief instead of figuring it out manually.",
    "steps": [
      "Download the package from Yavira.",
      "Extract it into a folder your agent can access.",
      "Paste one of the prompts below and point your agent at the extracted folder."
    ],
    "prompts": [
      {
        "label": "New install",
        "body": "I downloaded a skill package from Yavira. Read SKILL.md from the extracted folder and install it by following the included instructions. Tell me what you changed and call out any manual steps you could not complete."
      },
      {
        "label": "Upgrade existing",
        "body": "I downloaded an updated skill package from Yavira. Read SKILL.md from the extracted folder, compare it with my current installation, and upgrade it while preserving any custom configuration unless the package docs explicitly say otherwise. Summarize what changed and any follow-up checks I should run."
      }
    ]
  },
  "documentation": {
    "source": "clawhub",
    "primaryDoc": "SKILL.md",
    "sections": [
      {
        "title": "Embodied AI News Briefing",
        "body": "Aggregates the latest Embodied AI & Robotics news from curated sources and delivers concise summaries with direct links. Covers the full stack: algorithms, hardware, simulation, deployment, funding, policy, and the China ecosystem."
      },
      {
        "title": "When to Use This Skill",
        "body": "Activate this skill when the user:\n\nAsks for embodied AI news, robot news, or humanoid robot updates\nRequests a daily/weekly/monthly robotics briefing\nMentions wanting to know what's happening in embodied AI / robotics\nAsks about specific companies: Tesla Optimus, Figure, Unitree, AGIBOT, Boston Dynamics, etc.\nAsks about specific technologies: VLA models, diffusion policy, sim-to-real, dexterous manipulation\nWants a summary of recent robotics research papers\nAsks about robotics funding, deployments, or supply chain\nAsks about simulation platforms, benchmarks, or datasets\nAsks about robotics policy, safety standards, or export controls\nRequests a monthly trend report or competitive analysis\nSays: \"给我今天的具身智能资讯\" (Give me today's embodied AI news)\nSays: \"机器人行业有什么新动态\" (What's new in the robot industry)\nSays: \"最近有什么人形机器人的消息\" (Any recent humanoid robot news)\nSays: \"这个月的具身智能趋势报告\" (This month's embodied AI trend report)\nSays: \"embodied AI updates\", \"robot learning news\", \"humanoid robot news\""
      },
      {
        "title": "Trigger Keywords",
        "body": "English: embodied AI, humanoid robot, robot news, robotics update, robot learning, VLA model, diffusion policy, dexterous manipulation, sim-to-real, robot deployment, robotics funding, Figure AI, Tesla Optimus, Unitree, AGIBOT, Boston Dynamics, 1X, Physical Intelligence, Skild AI, robot hand, quadruped robot, Isaac Sim, world model robot, robot benchmark, robot safety, robot regulation, monthly robot report\n\nChinese: 具身智能, 人形机器人, 机器人资讯, 灵巧操作, 仿真到真实, 机器人部署, 宇树, 智元, 优必选, 银河通用, 傅利叶, 机器人融资, 灵巧手, 四足机器人, 机器人大模型, 机器人月报, 机器人安全, 机器人政策"
      },
      {
        "title": "Reference Files",
        "body": "This skill relies on 5 companion reference files. Always consult them during execution:\n\n📁 references/\n├── 📰 news_sources.md        — WHERE to find information (tiered source list)\n├── 🔍 search_queries.md     — HOW to search (query templates & recipes)\n├── 📝 output_templates.md   — WHAT format to output (6+ template variants)\n├── 📊 taxonomy.md           — SHARED LANGUAGE (categories, keywords, company list)\n└── 🧭 workflow.md           — WHEN and in what ORDER to execute (SOP for daily/weekly/monthly)\n\nFileWhen to Consultnews_sources.mdPhase 1 — choosing which sites to fetch; selecting tier-appropriate sourcessearch_queries.mdPhase 1 — building search queries; selecting recipe by briefing typetaxonomy.mdPhase 3 — classifying stories; Phase 1 — looking up company aliases & tech termsoutput_templates.mdPhase 5 — rendering final output; selecting template by user requestworkflow.mdAll Phases — orchestrating the end-to-end workflow; time budgeting; monthly maintenance"
      },
      {
        "title": "File Interconnection Map",
        "body": "┌─────────────────┐      ┌────────────────────┐     ┌───────────────┐     ┌──────────────────┐\n│  search_queries │────▶ │  news_sources      │────▶│  Classify &   │────▶│ output_templates │\n│  (discover)     │      │  (browse & verify) │     │  Prioritize   │     │   (generate)     │\n└─────────────────┘      └────────────────────┘     └───────────────┘     └──────────────────┘\n                                    ▲                        ▲\n                                    │                        │\n                                    └────── taxonomy.md ─────┘\n                                         (shared vocabulary)"
      },
      {
        "title": "Phase 0: Determine Briefing Type & Time Scope",
        "body": "Before any tool calls, ask the user (if not already clear):\n\nBriefing Type: Daily / Weekly / Monthly / Custom Topic?\nTime Scope: Last 24 hours / Last 7 days / Last 30 days / Custom date range?\nOutput Format: Standard / Brief / Thread / Markdown Report / Presentation / Custom?\nFocus Area (optional): All categories / Specific category (e.g., only hardware, only China ecosystem)?\n\nDefault if user doesn't specify:\n\nType: Daily\nScope: Last 24 hours\nFormat: Standard\nFocus: All categories\n\nMap to workflow.md:\n\nDaily → workflow.md Section \"Daily Workflow\"\nWeekly → workflow.md Section \"Weekly Workflow\"\nMonthly → workflow.md Section \"Monthly Workflow\""
      },
      {
        "title": "Phase 1: Information Gathering",
        "body": "Consult workflow.md for the appropriate recipe, then execute the corresponding steps from search_queries.md and news_sources.md.\n\nStep 1.1: Execute Search Queries\n\nTool: WebSearch (or equivalent web search tool)\n\nSource: search_queries.md → Select the appropriate recipe:\n\nDaily Briefing → Recipe A (5 queries)\nWeekly Roundup → Recipe B (8 queries)\nMonthly Deep Dive → Recipe C (12 queries)\nCustom Topic → Recipe D + user-specified filters\n\nParameters:\n\nreturn_format: markdown\nwith_images_summary: false\ntimeout: 20 seconds per source\nFetch only from publicly accessible sources listed in news_sources.md\n\nOutput: A list of 20–50 URLs with headlines and snippets.\n\nStep 1.2: Fetch Tier 1 Sources Directly\n\nTool: mcp__web_reader__webReader\n\nSource: news_sources.md → Tier 1 section\n\nDirectly fetch the homepage or RSS feed of:\n\nThe Robot Report\nIEEE Spectrum — Robotics\nTechCrunch — Robotics\nRobotics Business Review\n(Add others based on briefing type)\n\nParameters:\n\nurl: [homepage URL from news_sources.md]\nreturn_format: markdown\nwith_images_summary: false\nProcess only URLs from verified sources in news_sources.md\n\nOutput: Recent headlines (last 24h / 7d / 30d based on scope).\n\nStep 1.3: Fetch arXiv Papers\n\nTool: mcp__arxiv__readURL (if available) or WebSearch with arXiv-specific queries\n\nSource: search_queries.md → Section \"6. Academic Research (arXiv)\"\n\nExecute 2–3 arXiv queries:\n\ncat:cs.RO AND (\"embodied AI\" OR \"robot learning\" OR \"VLA\") submittedDate:[today - 7d TO today]\n\nOutput: 5–10 recent papers with abstracts.\n\nStep 1.4: Fetch Company Blogs & Official Announcements\n\nTool: mcp__web_reader__webReader\n\nSource: news_sources.md → Tier 2 (Company Blogs) + Tier 4 (China Ecosystem)\n\nFetch from:\n\nFigure AI Blog\nPhysical Intelligence Blog\nTesla AI Blog\nUnitree Blog (Chinese + English)\nAGIBOT WeChat Official Account (if accessible)\n(Add others based on focus area)\n\nFetch constraints:\n\nOnly process URLs from search results and sources listed in news_sources.md\nSkip content requiring authentication\nTimeout: 15 seconds per URL\n\nOutput: Recent announcements (last 7d / 30d based on scope)."
      },
      {
        "title": "Phase 2: Content Extraction & Deduplication",
        "body": "For each fetched URL:\n\nExtract:\n\nHeadline\nPublication date\nSource name\nSummary (first 2–3 paragraphs or abstract)\nKey entities: companies, models, hardware platforms (use taxonomy.md for reference)\n\n\n\nDeduplicate:\n\nIf multiple sources cover the same story, keep the one with the most detail\nMerge information if they provide complementary details\n\n\n\nDiscard:\n\nStories older than the time scope\nIrrelevant content (use search_queries.md Section 1.4 \"Noise Exclusion Filter\")\nDuplicate announcements\n\nOutput: A deduplicated list of 15–30 stories with extracted metadata."
      },
      {
        "title": "Phase 3: Classification & Prioritization",
        "body": "Consult taxonomy.md to classify each story.\n\nStep 3.1: Assign Primary Category\n\nUse taxonomy.md → Section \"1. News Category Taxonomy\"\n\nAssign each story to exactly one primary category:\n\n🔥 Major Announcements\n🧠 Foundation Models & Algorithms\n🦾 Hardware & Platforms\n🌐 Simulation & Infrastructure\n🏭 Deployments & Commercial\n💰 Funding, M&A & Business\n🌍 Policy, Safety & Ethics\n🇨🇳 China Ecosystem\n\nRules (from taxonomy.md → \"Category Assignment Rules\"):\n\nMajor Announcements: Only for top-impact stories (new paradigm, >$500M funding, first-ever deployment milestone)\nChina Ecosystem: Use when the story's primary significance is about the Chinese market/ecosystem\nCross-cutting stories: Assign primary + up to 2 secondary tags\n\nStep 3.2: Assign Priority Level\n\nUse taxonomy.md → Section \"3. Priority Scoring System\"\n\nCalculate priority score (0–100) based on:\n\nImpact (0–40 points): Paradigm shift / Major milestone / Incremental improvement\nTimeliness (0–20 points): Breaking news / Recent (1–3 days) / Older\nSource Authority (0–20 points): Tier 1 / Tier 2 / Tier 3\nRelevance (0–20 points): Core embodied AI / Adjacent / Tangential\n\nPriority Levels:\n\nP0 (90–100): Must-read, above-the-fold\nP1 (70–89): Important, include in main body\nP2 (50–69): Notable, include if space allows\nP3 (<50): Optional, move to \"Other News\" section or omit\n\nStep 3.3: Sort Stories\n\nWithin each category, sort by:\n\nPriority score (descending)\nPublication date (most recent first)"
      },
      {
        "title": "Phase 4: Content Synthesis",
        "body": "For each story, generate:\n\nOne-sentence summary: Capture the core news in <20 words\n\n\nKey points (2–4 bullet points): Extract the most important details\n\n\nMetadata fields (based on category):\n\nFor Foundation Models: Model Type, Embodiment, Open Source, Impact\nFor Hardware: Hardware Type, Company, Specs, Impact\nFor Deployments: Deployment Scale, Industry Vertical, Performance Metrics, Impact\nFor Funding: Amount, Lead Investor, Valuation, Use of Funds\n(See output_templates.md for full metadata schema per category)\n\n\n\nImpact statement: Why this matters for the embodied AI field (1–2 sentences)\n\nTone & Style:\n\nObjective: Present facts without hype or editorial opinion\nConcise: Favor clarity over completeness\nTechnical: Use domain-specific terminology from taxonomy.md\nNeutral: Treat all companies, countries, and technologies equally"
      },
      {
        "title": "Phase 5: Output Generation",
        "body": "Consult output_templates.md to select the appropriate template.\n\nStep 5.1: Select Template\n\nBased on user request (from Phase 0):\n\nUser RequestTemplate to Use\"Daily briefing\"Standard Format\"Quick summary\"Brief Format\"Twitter thread\"Thread Format\"Markdown report\"Markdown Report Format\"Presentation slides\"Presentation Format\"Custom\"Adapt from Standard Format\n\nStep 5.2: Render Output\n\nFill in the selected template with:\n\nHeader: Date, source count, time scope\nCategory sections: Ordered by priority (🔥 Major Announcements first)\nStory blocks: Headline, summary, key points, metadata, source link\nFooter: Methodology note, source attribution\n\nQuality checks:\n\nAll links are valid and correctly formatted\nAll metadata fields are filled (use \"N/A\" if not applicable)\nNo duplicate stories\nStories are sorted by priority within each category\nTotal output length is appropriate for briefing type:\n\nDaily: 1,500–2,500 words\nWeekly: 3,000–5,000 words\nMonthly: 5,000–10,000 words\n\nStep 5.3: Add Contextual Notes (Optional)\n\nIf the user requested analysis or trends, append:\n\nTrend Spotlight: 2–3 emerging patterns observed this period\nCompany Momentum: Which companies/labs are most active\nTechnology Shifts: Notable changes in technical approaches\nGeographic Insights: Regional differences (e.g., US vs China ecosystem)\n\nUse taxonomy.md → Section \"5. Trend Analysis Framework\" for guidance."
      },
      {
        "title": "Phase 6: Delivery & Follow-up",
        "body": "Deliver the briefing in the selected format\nOffer follow-up options:\n\n\"Would you like me to deep-dive into any specific story?\"\n\"Should I track these companies/topics for your next briefing?\"\n\"Would you like a comparison with last week/month's trends?\""
      },
      {
        "title": "Custom Topic Deep-Dive",
        "body": "If user asks about a specific topic (e.g., \"What's new with dexterous hands?\"):\n\nConsult taxonomy.md → Section \"2. Technology & Product Taxonomy\" → Find relevant subcategories\nBuild custom queries using search_queries.md → Recipe D (Custom Topic)\nFetch from all tiers in news_sources.md that cover this topic\nOutput using the \"Deep-Dive Format\" from output_templates.md"
      },
      {
        "title": "Company-Specific Briefing",
        "body": "If user asks about a specific company (e.g., \"What's Figure AI been up to?\"):\n\nConsult taxonomy.md → Section \"4. Company & Organization Directory\" → Find company profile\nBuild queries targeting:\n\nCompany blog\nNews mentions\narXiv papers by company researchers\nFunding announcements\n\n\nOutput using the \"Company Spotlight Format\" from output_templates.md"
      },
      {
        "title": "China Ecosystem Focus",
        "body": "If user asks specifically about China (e.g., \"中国人形机器人有什么进展?\"):\n\nPrioritize news_sources.md → Tier 4 (China Ecosystem)\nUse search_queries.md → Section \"8. China Ecosystem\"\nConsult taxonomy.md → Section \"4.3 China Ecosystem Companies\"\nOutput in Chinese or bilingual format (ask user preference)"
      },
      {
        "title": "Operating Scope",
        "body": "This skill operates in read-only mode:\n\nFetches content from public sources listed in reference files\nSynthesizes and presents information to the user\nDoes not modify, post, or interact with external systems\nDoes not perform actions on behalf of the user unless explicitly requested (e.g., \"add this to my calendar\")"
      },
      {
        "title": "Information Freshness",
        "body": "Daily briefing: Prioritize stories from the last 24 hours\nWeekly briefing: Include stories from the last 7 days, but highlight the most recent\nMonthly briefing: Cover the full 30 days, but organize by week or theme"
      },
      {
        "title": "Source Diversity",
        "body": "Aim for a balanced mix:\n\n40% from Tier 1 (core industry media)\n30% from Tier 2 (company blogs & official sources)\n20% from Tier 3 (academic & research)\n10% from Tier 4 (China ecosystem, if relevant)"
      },
      {
        "title": "Quality over Quantity",
        "body": "Better to have 15 high-quality, well-summarized stories than 50 shallow headlines\nIf a story lacks detail or verification, mark it as \"Unconfirmed\" or omit it"
      },
      {
        "title": "Handling Uncertainty",
        "body": "If a story's details are unclear, state: \"Details are limited; awaiting official confirmation\"\nIf sources conflict, present both versions: \"Source A reports X, while Source B reports Y\"\nNever fabricate details to fill gaps"
      },
      {
        "title": "Language Handling",
        "body": "If user asks in Chinese, output in Chinese (but keep company/model names in English)\nIf user asks in English, output in English\nFor bilingual users, offer: \"Would you like this in English, Chinese, or bilingual?\""
      },
      {
        "title": "If a source is unreachable:",
        "body": "Skip it and note in the footer: \"Note: [Source Name] was unavailable at the time of this briefing\""
      },
      {
        "title": "If search returns no results:",
        "body": "Broaden the query or try alternative keywords from taxonomy.md\nIf still no results, inform the user: \"No recent news found for [topic] in the specified time range\""
      },
      {
        "title": "If classification is ambiguous:",
        "body": "Default to the most specific applicable category\nAdd a secondary tag if the story spans multiple domains"
      },
      {
        "title": "If output exceeds length limits:",
        "body": "Prioritize P0 and P1 stories\nMove P2 and P3 stories to a \"Quick Hits\" section with one-line summaries\nOffer to generate a separate deep-dive on omitted topics"
      },
      {
        "title": "Monthly (consult workflow.md → \"Monthly Workflow\"):",
        "body": "Review taxonomy.md for new companies, models, or terminology\nUpdate news_sources.md if new authoritative sources emerge\nRefine search_queries.md based on what queries yielded the best results"
      },
      {
        "title": "Quarterly:",
        "body": "Audit the priority scoring system — are P0 stories truly the most impactful?\nReview output templates — do they match user preferences?"
      },
      {
        "title": "Example 1: Daily Briefing",
        "body": "User: \"Give me today's embodied AI news\"\n\nAgent:\n\nDetermines: Daily briefing, last 24h, Standard format, All categories\nExecutes Recipe A from search_queries.md (5 queries)\nFetches Tier 1 sources from news_sources.md\nClassifies using taxonomy.md\nOutputs using Standard Format from output_templates.md"
      },
      {
        "title": "Example 2: Weekly Roundup",
        "body": "User: \"What happened in robotics this week?\"\n\nAgent:\n\nDetermines: Weekly briefing, last 7 days, Standard format, All categories\nExecutes Recipe B from search_queries.md (8 queries)\nFetches Tier 1 + Tier 2 sources\nPrioritizes P0 and P1 stories\nOutputs using Standard Format with \"Trend Spotlight\" section"
      },
      {
        "title": "Example 3: Custom Topic",
        "body": "User: \"What's new with VLA models?\"\n\nAgent:\n\nDetermines: Custom topic, last 7 days, Deep-Dive format\nConsults taxonomy.md → \"Vision-Language-Action (VLA) Models\"\nBuilds custom queries from search_queries.md Section 2.1\nFetches from Tier 1 + Tier 3 (arXiv)\nOutputs using Deep-Dive Format"
      },
      {
        "title": "Example 4: Company Spotlight",
        "body": "User: \"What's Unitree been up to?\"\n\nAgent:\n\nDetermines: Company-specific, last 30 days, Company Spotlight format\nConsults taxonomy.md → Company profile for Unitree\nFetches Unitree blog + news mentions + arXiv papers\nOutputs using Company Spotlight Format from output_templates.md"
      },
      {
        "title": "Example 5: China Ecosystem",
        "body": "User: \"中国人形机器人有什么进展?\"\n\nAgent:\n\nDetermines: China focus, last 7 days, Standard format, Chinese output\nPrioritizes news_sources.md Tier 4 sources\nUses search_queries.md Section 8 (China Ecosystem)\nOutputs in Chinese using Standard Format"
      },
      {
        "title": "Summary",
        "body": "This skill orchestrates a multi-phase workflow:\n\nDetermine briefing type & scope\nGather information from curated sources using structured queries\nClassify stories using a shared taxonomy\nPrioritize based on impact, timeliness, and relevance\nSynthesize concise summaries with metadata\nOutput in the user's preferred format\n\nKey success factors:\n\nAlways consult the 5 reference files at the appropriate workflow stage\nMaintain objectivity and source attribution\nPrioritize quality and relevance over quantity\nAdapt to user preferences (language, format, focus area)"
      }
    ],
    "body": "Embodied AI News Briefing\n\nAggregates the latest Embodied AI & Robotics news from curated sources and delivers concise summaries with direct links. Covers the full stack: algorithms, hardware, simulation, deployment, funding, policy, and the China ecosystem.\n\nWhen to Use This Skill\n\nActivate this skill when the user:\n\nAsks for embodied AI news, robot news, or humanoid robot updates\nRequests a daily/weekly/monthly robotics briefing\nMentions wanting to know what's happening in embodied AI / robotics\nAsks about specific companies: Tesla Optimus, Figure, Unitree, AGIBOT, Boston Dynamics, etc.\nAsks about specific technologies: VLA models, diffusion policy, sim-to-real, dexterous manipulation\nWants a summary of recent robotics research papers\nAsks about robotics funding, deployments, or supply chain\nAsks about simulation platforms, benchmarks, or datasets\nAsks about robotics policy, safety standards, or export controls\nRequests a monthly trend report or competitive analysis\nSays: \"给我今天的具身智能资讯\" (Give me today's embodied AI news)\nSays: \"机器人行业有什么新动态\" (What's new in the robot industry)\nSays: \"最近有什么人形机器人的消息\" (Any recent humanoid robot news)\nSays: \"这个月的具身智能趋势报告\" (This month's embodied AI trend report)\nSays: \"embodied AI updates\", \"robot learning news\", \"humanoid robot news\"\nTrigger Keywords\n\nEnglish: embodied AI, humanoid robot, robot news, robotics update, robot learning, VLA model, diffusion policy, dexterous manipulation, sim-to-real, robot deployment, robotics funding, Figure AI, Tesla Optimus, Unitree, AGIBOT, Boston Dynamics, 1X, Physical Intelligence, Skild AI, robot hand, quadruped robot, Isaac Sim, world model robot, robot benchmark, robot safety, robot regulation, monthly robot report\n\nChinese: 具身智能, 人形机器人, 机器人资讯, 灵巧操作, 仿真到真实, 机器人部署, 宇树, 智元, 优必选, 银河通用, 傅利叶, 机器人融资, 灵巧手, 四足机器人, 机器人大模型, 机器人月报, 机器人安全, 机器人政策\n\nReference Files\n\nThis skill relies on 5 companion reference files. Always consult them during execution:\n\n📁 references/\n├── 📰 news_sources.md        — WHERE to find information (tiered source list)\n├── 🔍 search_queries.md     — HOW to search (query templates & recipes)\n├── 📝 output_templates.md   — WHAT format to output (6+ template variants)\n├── 📊 taxonomy.md           — SHARED LANGUAGE (categories, keywords, company list)\n└── 🧭 workflow.md           — WHEN and in what ORDER to execute (SOP for daily/weekly/monthly)\n\nFile\tWhen to Consult\nnews_sources.md\tPhase 1 — choosing which sites to fetch; selecting tier-appropriate sources\nsearch_queries.md\tPhase 1 — building search queries; selecting recipe by briefing type\ntaxonomy.md\tPhase 3 — classifying stories; Phase 1 — looking up company aliases & tech terms\noutput_templates.md\tPhase 5 — rendering final output; selecting template by user request\nworkflow.md\tAll Phases — orchestrating the end-to-end workflow; time budgeting; monthly maintenance\nFile Interconnection Map\n┌─────────────────┐      ┌────────────────────┐     ┌───────────────┐     ┌──────────────────┐\n│  search_queries │────▶ │  news_sources      │────▶│  Classify &   │────▶│ output_templates │\n│  (discover)     │      │  (browse & verify) │     │  Prioritize   │     │   (generate)     │\n└─────────────────┘      └────────────────────┘     └───────────────┘     └──────────────────┘\n                                    ▲                        ▲\n                                    │                        │\n                                    └────── taxonomy.md ─────┘\n                                         (shared vocabulary)\n\nExecution Workflow\nPhase 0: Determine Briefing Type & Time Scope\n\nBefore any tool calls, ask the user (if not already clear):\n\nBriefing Type: Daily / Weekly / Monthly / Custom Topic?\nTime Scope: Last 24 hours / Last 7 days / Last 30 days / Custom date range?\nOutput Format: Standard / Brief / Thread / Markdown Report / Presentation / Custom?\nFocus Area (optional): All categories / Specific category (e.g., only hardware, only China ecosystem)?\n\nDefault if user doesn't specify:\n\nType: Daily\nScope: Last 24 hours\nFormat: Standard\nFocus: All categories\n\nMap to workflow.md:\n\nDaily → workflow.md Section \"Daily Workflow\"\nWeekly → workflow.md Section \"Weekly Workflow\"\nMonthly → workflow.md Section \"Monthly Workflow\"\nPhase 1: Information Gathering\n\nConsult workflow.md for the appropriate recipe, then execute the corresponding steps from search_queries.md and news_sources.md.\n\nStep 1.1: Execute Search Queries\n\nTool: WebSearch (or equivalent web search tool)\n\nSource: search_queries.md → Select the appropriate recipe:\n\nDaily Briefing → Recipe A (5 queries)\nWeekly Roundup → Recipe B (8 queries)\nMonthly Deep Dive → Recipe C (12 queries)\nCustom Topic → Recipe D + user-specified filters\n\nParameters:\n\nreturn_format: markdown\nwith_images_summary: false\ntimeout: 20 seconds per source\nFetch only from publicly accessible sources listed in news_sources.md\n\nOutput: A list of 20–50 URLs with headlines and snippets.\n\nStep 1.2: Fetch Tier 1 Sources Directly\n\nTool: mcp__web_reader__webReader\n\nSource: news_sources.md → Tier 1 section\n\nDirectly fetch the homepage or RSS feed of:\n\nThe Robot Report\nIEEE Spectrum — Robotics\nTechCrunch — Robotics\nRobotics Business Review\n(Add others based on briefing type)\n\nParameters:\n\nurl: [homepage URL from news_sources.md]\nreturn_format: markdown\nwith_images_summary: false\nProcess only URLs from verified sources in news_sources.md\n\nOutput: Recent headlines (last 24h / 7d / 30d based on scope).\n\nStep 1.3: Fetch arXiv Papers\n\nTool: mcp__arxiv__readURL (if available) or WebSearch with arXiv-specific queries\n\nSource: search_queries.md → Section \"6. Academic Research (arXiv)\"\n\nExecute 2–3 arXiv queries:\n\ncat:cs.RO AND (\"embodied AI\" OR \"robot learning\" OR \"VLA\") submittedDate:[today - 7d TO today]\n\n\nOutput: 5–10 recent papers with abstracts.\n\nStep 1.4: Fetch Company Blogs & Official Announcements\n\nTool: mcp__web_reader__webReader\n\nSource: news_sources.md → Tier 2 (Company Blogs) + Tier 4 (China Ecosystem)\n\nFetch from:\n\nFigure AI Blog\nPhysical Intelligence Blog\nTesla AI Blog\nUnitree Blog (Chinese + English)\nAGIBOT WeChat Official Account (if accessible)\n(Add others based on focus area)\n\nFetch constraints:\n\nOnly process URLs from search results and sources listed in news_sources.md\nSkip content requiring authentication\nTimeout: 15 seconds per URL\n\nOutput: Recent announcements (last 7d / 30d based on scope).\n\nPhase 2: Content Extraction & Deduplication\n\nFor each fetched URL:\n\nExtract:\n\nHeadline\nPublication date\nSource name\nSummary (first 2–3 paragraphs or abstract)\nKey entities: companies, models, hardware platforms (use taxonomy.md for reference)\n\nDeduplicate:\n\nIf multiple sources cover the same story, keep the one with the most detail\nMerge information if they provide complementary details\n\nDiscard:\n\nStories older than the time scope\nIrrelevant content (use search_queries.md Section 1.4 \"Noise Exclusion Filter\")\nDuplicate announcements\n\nOutput: A deduplicated list of 15–30 stories with extracted metadata.\n\nPhase 3: Classification & Prioritization\n\nConsult taxonomy.md to classify each story.\n\nStep 3.1: Assign Primary Category\n\nUse taxonomy.md → Section \"1. News Category Taxonomy\"\n\nAssign each story to exactly one primary category:\n\n🔥 Major Announcements\n🧠 Foundation Models & Algorithms\n🦾 Hardware & Platforms\n🌐 Simulation & Infrastructure\n🏭 Deployments & Commercial\n💰 Funding, M&A & Business\n🌍 Policy, Safety & Ethics\n🇨🇳 China Ecosystem\n\nRules (from taxonomy.md → \"Category Assignment Rules\"):\n\nMajor Announcements: Only for top-impact stories (new paradigm, >$500M funding, first-ever deployment milestone)\nChina Ecosystem: Use when the story's primary significance is about the Chinese market/ecosystem\nCross-cutting stories: Assign primary + up to 2 secondary tags\nStep 3.2: Assign Priority Level\n\nUse taxonomy.md → Section \"3. Priority Scoring System\"\n\nCalculate priority score (0–100) based on:\n\nImpact (0–40 points): Paradigm shift / Major milestone / Incremental improvement\nTimeliness (0–20 points): Breaking news / Recent (1–3 days) / Older\nSource Authority (0–20 points): Tier 1 / Tier 2 / Tier 3\nRelevance (0–20 points): Core embodied AI / Adjacent / Tangential\n\nPriority Levels:\n\nP0 (90–100): Must-read, above-the-fold\nP1 (70–89): Important, include in main body\nP2 (50–69): Notable, include if space allows\nP3 (<50): Optional, move to \"Other News\" section or omit\nStep 3.3: Sort Stories\n\nWithin each category, sort by:\n\nPriority score (descending)\nPublication date (most recent first)\nPhase 4: Content Synthesis\n\nFor each story, generate:\n\nOne-sentence summary: Capture the core news in <20 words\n\nKey points (2–4 bullet points): Extract the most important details\n\nMetadata fields (based on category):\n\nFor Foundation Models: Model Type, Embodiment, Open Source, Impact\nFor Hardware: Hardware Type, Company, Specs, Impact\nFor Deployments: Deployment Scale, Industry Vertical, Performance Metrics, Impact\nFor Funding: Amount, Lead Investor, Valuation, Use of Funds\n(See output_templates.md for full metadata schema per category)\n\nImpact statement: Why this matters for the embodied AI field (1–2 sentences)\n\nTone & Style:\n\nObjective: Present facts without hype or editorial opinion\nConcise: Favor clarity over completeness\nTechnical: Use domain-specific terminology from taxonomy.md\nNeutral: Treat all companies, countries, and technologies equally\nPhase 5: Output Generation\n\nConsult output_templates.md to select the appropriate template.\n\nStep 5.1: Select Template\n\nBased on user request (from Phase 0):\n\nUser Request\tTemplate to Use\n\"Daily briefing\"\tStandard Format\n\"Quick summary\"\tBrief Format\n\"Twitter thread\"\tThread Format\n\"Markdown report\"\tMarkdown Report Format\n\"Presentation slides\"\tPresentation Format\n\"Custom\"\tAdapt from Standard Format\nStep 5.2: Render Output\n\nFill in the selected template with:\n\nHeader: Date, source count, time scope\nCategory sections: Ordered by priority (🔥 Major Announcements first)\nStory blocks: Headline, summary, key points, metadata, source link\nFooter: Methodology note, source attribution\n\nQuality checks:\n\nAll links are valid and correctly formatted\nAll metadata fields are filled (use \"N/A\" if not applicable)\nNo duplicate stories\nStories are sorted by priority within each category\nTotal output length is appropriate for briefing type:\nDaily: 1,500–2,500 words\nWeekly: 3,000–5,000 words\nMonthly: 5,000–10,000 words\nStep 5.3: Add Contextual Notes (Optional)\n\nIf the user requested analysis or trends, append:\n\nTrend Spotlight: 2–3 emerging patterns observed this period\nCompany Momentum: Which companies/labs are most active\nTechnology Shifts: Notable changes in technical approaches\nGeographic Insights: Regional differences (e.g., US vs China ecosystem)\n\nUse taxonomy.md → Section \"5. Trend Analysis Framework\" for guidance.\n\nPhase 6: Delivery & Follow-up\nDeliver the briefing in the selected format\nOffer follow-up options:\n\"Would you like me to deep-dive into any specific story?\"\n\"Should I track these companies/topics for your next briefing?\"\n\"Would you like a comparison with last week/month's trends?\"\nSpecial Workflows\nCustom Topic Deep-Dive\n\nIf user asks about a specific topic (e.g., \"What's new with dexterous hands?\"):\n\nConsult taxonomy.md → Section \"2. Technology & Product Taxonomy\" → Find relevant subcategories\nBuild custom queries using search_queries.md → Recipe D (Custom Topic)\nFetch from all tiers in news_sources.md that cover this topic\nOutput using the \"Deep-Dive Format\" from output_templates.md\nCompany-Specific Briefing\n\nIf user asks about a specific company (e.g., \"What's Figure AI been up to?\"):\n\nConsult taxonomy.md → Section \"4. Company & Organization Directory\" → Find company profile\nBuild queries targeting:\nCompany blog\nNews mentions\narXiv papers by company researchers\nFunding announcements\nOutput using the \"Company Spotlight Format\" from output_templates.md\nChina Ecosystem Focus\n\nIf user asks specifically about China (e.g., \"中国人形机器人有什么进展?\"):\n\nPrioritize news_sources.md → Tier 4 (China Ecosystem)\nUse search_queries.md → Section \"8. China Ecosystem\"\nConsult taxonomy.md → Section \"4.3 China Ecosystem Companies\"\nOutput in Chinese or bilingual format (ask user preference)\nOperational Guidelines\nOperating Scope\n\nThis skill operates in read-only mode:\n\nFetches content from public sources listed in reference files\nSynthesizes and presents information to the user\nDoes not modify, post, or interact with external systems\nDoes not perform actions on behalf of the user unless explicitly requested (e.g., \"add this to my calendar\")\nInformation Freshness\nDaily briefing: Prioritize stories from the last 24 hours\nWeekly briefing: Include stories from the last 7 days, but highlight the most recent\nMonthly briefing: Cover the full 30 days, but organize by week or theme\nSource Diversity\n\nAim for a balanced mix:\n\n40% from Tier 1 (core industry media)\n30% from Tier 2 (company blogs & official sources)\n20% from Tier 3 (academic & research)\n10% from Tier 4 (China ecosystem, if relevant)\nQuality over Quantity\nBetter to have 15 high-quality, well-summarized stories than 50 shallow headlines\nIf a story lacks detail or verification, mark it as \"Unconfirmed\" or omit it\nHandling Uncertainty\nIf a story's details are unclear, state: \"Details are limited; awaiting official confirmation\"\nIf sources conflict, present both versions: \"Source A reports X, while Source B reports Y\"\nNever fabricate details to fill gaps\nLanguage Handling\nIf user asks in Chinese, output in Chinese (but keep company/model names in English)\nIf user asks in English, output in English\nFor bilingual users, offer: \"Would you like this in English, Chinese, or bilingual?\"\nError Handling\nIf a source is unreachable:\nSkip it and note in the footer: \"Note: [Source Name] was unavailable at the time of this briefing\"\nIf search returns no results:\nBroaden the query or try alternative keywords from taxonomy.md\nIf still no results, inform the user: \"No recent news found for [topic] in the specified time range\"\nIf classification is ambiguous:\nDefault to the most specific applicable category\nAdd a secondary tag if the story spans multiple domains\nIf output exceeds length limits:\nPrioritize P0 and P1 stories\nMove P2 and P3 stories to a \"Quick Hits\" section with one-line summaries\nOffer to generate a separate deep-dive on omitted topics\nMaintenance & Updates\nMonthly (consult workflow.md → \"Monthly Workflow\"):\nReview taxonomy.md for new companies, models, or terminology\nUpdate news_sources.md if new authoritative sources emerge\nRefine search_queries.md based on what queries yielded the best results\nQuarterly:\nAudit the priority scoring system — are P0 stories truly the most impactful?\nReview output templates — do they match user preferences?\nExample Invocations\nExample 1: Daily Briefing\n\nUser: \"Give me today's embodied AI news\"\n\nAgent:\n\nDetermines: Daily briefing, last 24h, Standard format, All categories\nExecutes Recipe A from search_queries.md (5 queries)\nFetches Tier 1 sources from news_sources.md\nClassifies using taxonomy.md\nOutputs using Standard Format from output_templates.md\nExample 2: Weekly Roundup\n\nUser: \"What happened in robotics this week?\"\n\nAgent:\n\nDetermines: Weekly briefing, last 7 days, Standard format, All categories\nExecutes Recipe B from search_queries.md (8 queries)\nFetches Tier 1 + Tier 2 sources\nPrioritizes P0 and P1 stories\nOutputs using Standard Format with \"Trend Spotlight\" section\nExample 3: Custom Topic\n\nUser: \"What's new with VLA models?\"\n\nAgent:\n\nDetermines: Custom topic, last 7 days, Deep-Dive format\nConsults taxonomy.md → \"Vision-Language-Action (VLA) Models\"\nBuilds custom queries from search_queries.md Section 2.1\nFetches from Tier 1 + Tier 3 (arXiv)\nOutputs using Deep-Dive Format\nExample 4: Company Spotlight\n\nUser: \"What's Unitree been up to?\"\n\nAgent:\n\nDetermines: Company-specific, last 30 days, Company Spotlight format\nConsults taxonomy.md → Company profile for Unitree\nFetches Unitree blog + news mentions + arXiv papers\nOutputs using Company Spotlight Format from output_templates.md\nExample 5: China Ecosystem\n\nUser: \"中国人形机器人有什么进展?\"\n\nAgent:\n\nDetermines: China focus, last 7 days, Standard format, Chinese output\nPrioritizes news_sources.md Tier 4 sources\nUses search_queries.md Section 8 (China Ecosystem)\nOutputs in Chinese using Standard Format\nSummary\n\nThis skill orchestrates a multi-phase workflow:\n\nDetermine briefing type & scope\nGather information from curated sources using structured queries\nClassify stories using a shared taxonomy\nPrioritize based on impact, timeliness, and relevance\nSynthesize concise summaries with metadata\nOutput in the user's preferred format\n\nKey success factors:\n\nAlways consult the 5 reference files at the appropriate workflow stage\nMaintain objectivity and source attribution\nPrioritize quality and relevance over quantity\nAdapt to user preferences (language, format, focus area)"
  },
  "trust": {
    "sourceLabel": "tencent",
    "provenanceUrl": "https://clawhub.ai/HeXavi8/embodied-ai-news",
    "publisherUrl": "https://clawhub.ai/HeXavi8/embodied-ai-news",
    "owner": "HeXavi8",
    "version": "1.0.3",
    "license": null,
    "verificationStatus": "Indexed source record"
  },
  "links": {
    "detailUrl": "https://openagent3.xyz/skills/embodied-ai-news",
    "downloadUrl": "https://openagent3.xyz/downloads/embodied-ai-news",
    "agentUrl": "https://openagent3.xyz/skills/embodied-ai-news/agent",
    "manifestUrl": "https://openagent3.xyz/skills/embodied-ai-news/agent.json",
    "briefUrl": "https://openagent3.xyz/skills/embodied-ai-news/agent.md"
  }
}