{
  "schemaVersion": "1.0",
  "item": {
    "slug": "opusflame-deep-research",
    "name": "OpusFlame Deep Research",
    "source": "tencent",
    "type": "skill",
    "category": "效率提升",
    "sourceUrl": "https://clawhub.ai/LeadingOT/opusflame-deep-research",
    "canonicalUrl": "https://clawhub.ai/LeadingOT/opusflame-deep-research",
    "targetPlatform": "OpenClaw"
  },
  "install": {
    "downloadMode": "redirect",
    "downloadUrl": "/downloads/opusflame-deep-research",
    "sourceDownloadUrl": "https://wry-manatee-359.convex.site/api/v1/download?slug=opusflame-deep-research",
    "sourcePlatform": "tencent",
    "targetPlatform": "OpenClaw",
    "installMethod": "Manual import",
    "extraction": "Extract archive",
    "prerequisites": [
      "OpenClaw"
    ],
    "packageFormat": "ZIP package",
    "includedAssets": [
      "SKILL.md",
      "references/financial-research.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",
      "status": "healthy",
      "reason": "direct_download_ok",
      "recommendedAction": "download",
      "checkedAt": "2026-04-30T16:55:25.780Z",
      "expiresAt": "2026-05-07T16:55:25.780Z",
      "httpStatus": 200,
      "finalUrl": "https://wry-manatee-359.convex.site/api/v1/download?slug=network",
      "contentType": "application/zip",
      "probeMethod": "head",
      "details": {
        "probeUrl": "https://wry-manatee-359.convex.site/api/v1/download?slug=network",
        "contentDisposition": "attachment; filename=\"network-1.0.0.zip\"",
        "redirectLocation": null,
        "bodySnippet": null
      },
      "scope": "source",
      "summary": "Source download looks usable.",
      "detail": "Yavira can redirect you to the upstream package for this source.",
      "primaryActionLabel": "Download for OpenClaw",
      "primaryActionHref": "/downloads/opusflame-deep-research"
    },
    "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/opusflame-deep-research",
    "agentPageUrl": "https://openagent3.xyz/skills/opusflame-deep-research/agent",
    "manifestUrl": "https://openagent3.xyz/skills/opusflame-deep-research/agent.json",
    "briefUrl": "https://openagent3.xyz/skills/opusflame-deep-research/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": "Deep Research (Multi-Model + Framework-Driven)",
        "body": "Autonomous research system that runs 4 AI models in parallel, each applying relevant analytical frameworks, then cross-validates and merges findings into a comprehensive cited report."
      },
      {
        "title": "Architecture",
        "body": "User Question\n     │\n     ▼\n┌─ Phase 0: Framework Selection ─┐\n│  Identify best-practice         │\n│  framework(s) for this question │\n└────────────┬────────────────────┘\n             │\n     ┌───────┼───────┐───────┐\n     ▼       ▼       ▼       ▼\n  Gemini    o3     Opus   MiniMax\n  2.5 Pro         4       M2.5\n  (search  (deep  (nuance (China/\n  heavy)   logic) +balance)alt view)\n     │       │       │       │\n     └───────┼───────┘───────┘\n             ▼\n      Phase 5: Merge & Cross-Validate\n             │\n             ▼\n       Final Report (PDF)"
      },
      {
        "title": "Phase 0: Framework Selection (MANDATORY — before any research)",
        "body": "Before researching, ask: \"Is there a best-practice framework for answering this type of question?\""
      },
      {
        "title": "Framework Lookup Table",
        "body": "Question TypeFrameworks to ApplyCompetitive strategyPorter's Five Forces, 7 Powers (Helmer), Schwerpunkt/High Ground (Packy), SWOTMarket entry / sizingTAM/SAM/SOM, Blue Ocean Strategy, Jobs-to-be-DoneBusiness model evaluationBusiness Model Canvas, Unit Economics, Ramp vs Route test (point solution vs platform?)Investment / valuationDCF, Comparable Analysis, Venture method, Power Law thesisProduct strategyJTBD, Kano Model, Value Prop Canvas, Hook ModelGrowth / GTMAARRR Pirate Metrics, Bullseye Framework, STP (Segmentation-Targeting-Positioning)Technology assessmentGartner Hype Cycle, Wardley Maps, Build vs Buy matrixRisk analysisPre-Mortem, FMEA, Scenario PlanningOrganizational / opsOKR analysis, RACI, Theory of ConstraintsPricingVan Westendorp, Conjoint, Value-based pricing frameworkIndustry analysisValue Chain Analysis, Industry Lifecycle, Winner-Takes-More thesisPerson / hiringTrack Record Analysis, Reference Triangle, Founder-Market Fit\n\nIf a framework applies:\n\nInclude it in the prompt to each model\nStructure the model's analysis around the framework's components\nThe final report should explicitly reference which framework(s) were used and why\n\nIf no standard framework applies:\n\nState \"No standard framework identified — using first-principles analysis\"\nEach model reasons from first principles with explicit assumptions stated"
      },
      {
        "title": "Phase 1: Decompose (30s)",
        "body": "Break the topic into 5-8 research sub-questions. Think like an investigative journalist:\n\nWhat are the key facts?\nWhat are different perspectives/sources?\nWhat's the timeline/history?\nWhat data/evidence exists?\nWhat are the unknowns or controversies?"
      },
      {
        "title": "Phase 2: Spawn 4 Model Agents (Parallel)",
        "body": "Spawn 4 sub-agents using sessions_spawn, each with a different model:\n\nModel 1: gemini       (google/gemini-2.5-pro)  — Search-heavy, broad coverage\nModel 2: o3           (openai/o3)              — Deep logical reasoning, contrarian\nModel 3: opus         (anthropic/claude-opus-4-6) — Nuanced, balanced synthesis\nModel 4: minimax      (minimax/MiniMax-M2.5)   — Alternative perspectives, China/grey-area"
      },
      {
        "title": "Prompt Template for Each Model",
        "body": "## Research Task\n[Topic]\n\n## Framework\nYou MUST structure your analysis using: [Framework Name]\nApply each component of the framework systematically to the topic.\nIf data is missing for a component, note it explicitly.\n\n## Sub-Questions\n[List of 5-8 sub-questions]\n\n## Instructions\n1. Use web_search extensively (minimum 10 unique searches)\n2. Use web_fetch to read full articles for key sources\n3. Cross-reference claims across 2+ sources\n4. Structure findings around the framework components\n5. Flag disagreements, unknowns, and low-confidence claims\n6. Minimum 15 unique source URLs\n7. Output format: markdown with inline citations [1][2]...\n8. End with a Sources section listing all URLs\n\n## Quality Rules\n- Every factual claim needs a source\n- Prefer primary sources (filings, official reports) over secondary\n- Note source freshness — flag anything >6 months old\n- Include opposing viewpoints\n- State confidence level (high/medium/low) for key conclusions"
      },
      {
        "title": "Model-Specific Instructions",
        "body": "Gemini: \"You are the primary search engine. Cast the widest net. Find obscure sources others would miss. Prioritize data and numbers.\"\no3: \"You are the deep reasoner. Challenge assumptions. Look for logical flaws in conventional wisdom. Apply the framework with maximum rigor. If the consensus is wrong, explain why.\"\nOpus: \"You are the synthesizer. Balance multiple perspectives fairly. Identify nuance others miss. Connect dots across disciplines.\"\nMiniMax: \"You are the alternative perspective agent. Consider non-Western viewpoints, grey areas, unconventional strategies. What would a Chinese entrepreneur or contrarian investor do differently?\""
      },
      {
        "title": "Phase 3: Wait for Completion",
        "body": "All 4 models run in parallel via sessions_spawn with mode=\"run\". Do NOT poll in a loop — they auto-announce when done."
      },
      {
        "title": "Phase 4: Collect Individual Reports",
        "body": "Save each model's output:\n\nmemory/research/[topic]-gemini-[date].md\nmemory/research/[topic]-o3-[date].md\nmemory/research/[topic]-opus-[date].md\nmemory/research/[topic]-minimax-[date].md"
      },
      {
        "title": "Phase 5: Cross-Validate & Merge",
        "body": "This is the most critical phase. The primary agent (you) must:"
      },
      {
        "title": "5a. Agreement Matrix",
        "body": "Create a matrix of key claims and which models agree/disagree:\n\n| Claim | Gemini | o3 | Opus | MiniMax | Confidence |\n|-------|--------|----|----|---------|------------|\n| [claim 1] | ✅ | ✅ | ✅ | ❌ | High (3/4) |\n| [claim 2] | ✅ | ❌ | ✅ | ✅ | High (3/4) |\n| [claim 3] | ✅ | ✅ | ❓ | ❓ | Medium (2/4) |"
      },
      {
        "title": "5b. Conflict Resolution",
        "body": "For each disagreement:\n\nIdentify the root cause (different data? different logic? different framework interpretation?)\nCheck which model has the stronger source\nIf genuinely uncertain, present both sides in the final report"
      },
      {
        "title": "5c. Framework Synthesis",
        "body": "Map findings back to the framework structure\nEnsure every framework component has been addressed\nNote which components had strong consensus vs. disagreement"
      },
      {
        "title": "5d. Error Catching",
        "body": "From experience, models commonly get wrong:\n\nPlatform-specific limits (posting frequency, API limits)\nPricing (especially for niche tools — often 10-30x off)\nRegulatory details\nRecency of data\n\nVerify any quantitative claim that only one model makes."
      },
      {
        "title": "Phase 6: Final Report",
        "body": "# [Topic] — Deep Research Report\n\n**Framework Used**: [Name] — [why this framework]\n**Models**: Gemini 2.5 Pro, o3, Opus 4, MiniMax M2.5\n**Date**: [date]\n**Total Searches**: [count across all models]\n\n## Executive Summary\n3-5 sentence overview. Note consensus level.\n\n## Framework Analysis\n\n### [Framework Component 1]\nAnalysis with model consensus noted. [1][2]\n\n### [Framework Component 2]\n...\n\n## Key Findings (Beyond Framework)\nDiscoveries that don't fit neatly into the framework.\n\n## Model Disagreements\nWhere models diverged and why.\n\n## Agreement Matrix\n[The table from 5a]\n\n## Data & Evidence\nTables, numbers, comparisons.\n\n## Risks / Unknowns\nWhat we couldn't confirm. Low-confidence areas.\n\n## Conclusion & Recommendations\nActionable takeaways ranked by confidence.\n\n## Sources\n[1] Title — URL\n[2] ..."
      },
      {
        "title": "Phase 7: Deliver",
        "body": "Save final report to memory/research/[topic]-终极版-[date].md\nGenerate PDF via pymupdf and save to ~/.openclaw/media/outbound/\nSend PDF to user via message tool"
      },
      {
        "title": "Quality Standards",
        "body": "Minimum sources: 15 unique URLs per model (60+ total across 4 models)\nSource diversity: No more than 3 citations from same domain per model\nFreshness: Prefer sources < 6 months old; flag older data\nCross-validation: Key claims must appear in 2+ models' findings\nFramework compliance: Every framework component must be addressed\nConfidence scoring: High (3-4 models agree + strong sources), Medium (2 models or weak sources), Low (1 model or no source)\nNo hallucination: Every factual claim must have a source"
      },
      {
        "title": "Financial / Stock Research",
        "body": "Frameworks: DCF, Comparable Analysis, Power Law\nCheck SEC/regulatory filings, earnings transcripts\nInclude key metrics (revenue, margins, P/E, debt)\nSee references/financial-research.md"
      },
      {
        "title": "Market / Industry Research",
        "body": "Frameworks: Porter's Five Forces, TAM/SAM/SOM, 7 Powers\nCompetitive landscape, key players, market share\nApply Winner-Takes-More thesis where relevant"
      },
      {
        "title": "Strategy / Business Model",
        "body": "Frameworks: Schwerpunkt/High Ground, Business Model Canvas, JTBD\nIdentify the constraint, the scarce asset, expansion path\nCompare to historical precedents (Rockefeller, Ramp, etc.)"
      },
      {
        "title": "Technical / Product Research",
        "body": "Frameworks: Wardley Maps, Build vs Buy, Gartner Hype Cycle\nArchitecture, benchmarks, alternatives matrix\nCommunity sentiment (GitHub, HN, Reddit)"
      }
    ],
    "body": "Deep Research (Multi-Model + Framework-Driven)\n\nAutonomous research system that runs 4 AI models in parallel, each applying relevant analytical frameworks, then cross-validates and merges findings into a comprehensive cited report.\n\nArchitecture\nUser Question\n     │\n     ▼\n┌─ Phase 0: Framework Selection ─┐\n│  Identify best-practice         │\n│  framework(s) for this question │\n└────────────┬────────────────────┘\n             │\n     ┌───────┼───────┐───────┐\n     ▼       ▼       ▼       ▼\n  Gemini    o3     Opus   MiniMax\n  2.5 Pro         4       M2.5\n  (search  (deep  (nuance (China/\n  heavy)   logic) +balance)alt view)\n     │       │       │       │\n     └───────┼───────┘───────┘\n             ▼\n      Phase 5: Merge & Cross-Validate\n             │\n             ▼\n       Final Report (PDF)\n\nPhase 0: Framework Selection (MANDATORY — before any research)\n\nBefore researching, ask: \"Is there a best-practice framework for answering this type of question?\"\n\nFramework Lookup Table\nQuestion Type\tFrameworks to Apply\nCompetitive strategy\tPorter's Five Forces, 7 Powers (Helmer), Schwerpunkt/High Ground (Packy), SWOT\nMarket entry / sizing\tTAM/SAM/SOM, Blue Ocean Strategy, Jobs-to-be-Done\nBusiness model evaluation\tBusiness Model Canvas, Unit Economics, Ramp vs Route test (point solution vs platform?)\nInvestment / valuation\tDCF, Comparable Analysis, Venture method, Power Law thesis\nProduct strategy\tJTBD, Kano Model, Value Prop Canvas, Hook Model\nGrowth / GTM\tAARRR Pirate Metrics, Bullseye Framework, STP (Segmentation-Targeting-Positioning)\nTechnology assessment\tGartner Hype Cycle, Wardley Maps, Build vs Buy matrix\nRisk analysis\tPre-Mortem, FMEA, Scenario Planning\nOrganizational / ops\tOKR analysis, RACI, Theory of Constraints\nPricing\tVan Westendorp, Conjoint, Value-based pricing framework\nIndustry analysis\tValue Chain Analysis, Industry Lifecycle, Winner-Takes-More thesis\nPerson / hiring\tTrack Record Analysis, Reference Triangle, Founder-Market Fit\n\nIf a framework applies:\n\nInclude it in the prompt to each model\nStructure the model's analysis around the framework's components\nThe final report should explicitly reference which framework(s) were used and why\n\nIf no standard framework applies:\n\nState \"No standard framework identified — using first-principles analysis\"\nEach model reasons from first principles with explicit assumptions stated\nPhase 1: Decompose (30s)\n\nBreak the topic into 5-8 research sub-questions. Think like an investigative journalist:\n\nWhat are the key facts?\nWhat are different perspectives/sources?\nWhat's the timeline/history?\nWhat data/evidence exists?\nWhat are the unknowns or controversies?\nPhase 2: Spawn 4 Model Agents (Parallel)\n\nSpawn 4 sub-agents using sessions_spawn, each with a different model:\n\nModel 1: gemini       (google/gemini-2.5-pro)  — Search-heavy, broad coverage\nModel 2: o3           (openai/o3)              — Deep logical reasoning, contrarian\nModel 3: opus         (anthropic/claude-opus-4-6) — Nuanced, balanced synthesis\nModel 4: minimax      (minimax/MiniMax-M2.5)   — Alternative perspectives, China/grey-area\n\nPrompt Template for Each Model\n## Research Task\n[Topic]\n\n## Framework\nYou MUST structure your analysis using: [Framework Name]\nApply each component of the framework systematically to the topic.\nIf data is missing for a component, note it explicitly.\n\n## Sub-Questions\n[List of 5-8 sub-questions]\n\n## Instructions\n1. Use web_search extensively (minimum 10 unique searches)\n2. Use web_fetch to read full articles for key sources\n3. Cross-reference claims across 2+ sources\n4. Structure findings around the framework components\n5. Flag disagreements, unknowns, and low-confidence claims\n6. Minimum 15 unique source URLs\n7. Output format: markdown with inline citations [1][2]...\n8. End with a Sources section listing all URLs\n\n## Quality Rules\n- Every factual claim needs a source\n- Prefer primary sources (filings, official reports) over secondary\n- Note source freshness — flag anything >6 months old\n- Include opposing viewpoints\n- State confidence level (high/medium/low) for key conclusions\n\nModel-Specific Instructions\nGemini: \"You are the primary search engine. Cast the widest net. Find obscure sources others would miss. Prioritize data and numbers.\"\no3: \"You are the deep reasoner. Challenge assumptions. Look for logical flaws in conventional wisdom. Apply the framework with maximum rigor. If the consensus is wrong, explain why.\"\nOpus: \"You are the synthesizer. Balance multiple perspectives fairly. Identify nuance others miss. Connect dots across disciplines.\"\nMiniMax: \"You are the alternative perspective agent. Consider non-Western viewpoints, grey areas, unconventional strategies. What would a Chinese entrepreneur or contrarian investor do differently?\"\nPhase 3: Wait for Completion\n\nAll 4 models run in parallel via sessions_spawn with mode=\"run\". Do NOT poll in a loop — they auto-announce when done.\n\nPhase 4: Collect Individual Reports\n\nSave each model's output:\n\nmemory/research/[topic]-gemini-[date].md\nmemory/research/[topic]-o3-[date].md\nmemory/research/[topic]-opus-[date].md\nmemory/research/[topic]-minimax-[date].md\n\nPhase 5: Cross-Validate & Merge\n\nThis is the most critical phase. The primary agent (you) must:\n\n5a. Agreement Matrix\n\nCreate a matrix of key claims and which models agree/disagree:\n\n| Claim | Gemini | o3 | Opus | MiniMax | Confidence |\n|-------|--------|----|----|---------|------------|\n| [claim 1] | ✅ | ✅ | ✅ | ❌ | High (3/4) |\n| [claim 2] | ✅ | ❌ | ✅ | ✅ | High (3/4) |\n| [claim 3] | ✅ | ✅ | ❓ | ❓ | Medium (2/4) |\n\n5b. Conflict Resolution\n\nFor each disagreement:\n\nIdentify the root cause (different data? different logic? different framework interpretation?)\nCheck which model has the stronger source\nIf genuinely uncertain, present both sides in the final report\n5c. Framework Synthesis\nMap findings back to the framework structure\nEnsure every framework component has been addressed\nNote which components had strong consensus vs. disagreement\n5d. Error Catching\n\nFrom experience, models commonly get wrong:\n\nPlatform-specific limits (posting frequency, API limits)\nPricing (especially for niche tools — often 10-30x off)\nRegulatory details\nRecency of data\n\nVerify any quantitative claim that only one model makes.\n\nPhase 6: Final Report\n# [Topic] — Deep Research Report\n\n**Framework Used**: [Name] — [why this framework]\n**Models**: Gemini 2.5 Pro, o3, Opus 4, MiniMax M2.5\n**Date**: [date]\n**Total Searches**: [count across all models]\n\n## Executive Summary\n3-5 sentence overview. Note consensus level.\n\n## Framework Analysis\n\n### [Framework Component 1]\nAnalysis with model consensus noted. [1][2]\n\n### [Framework Component 2]\n...\n\n## Key Findings (Beyond Framework)\nDiscoveries that don't fit neatly into the framework.\n\n## Model Disagreements\nWhere models diverged and why.\n\n## Agreement Matrix\n[The table from 5a]\n\n## Data & Evidence\nTables, numbers, comparisons.\n\n## Risks / Unknowns\nWhat we couldn't confirm. Low-confidence areas.\n\n## Conclusion & Recommendations\nActionable takeaways ranked by confidence.\n\n## Sources\n[1] Title — URL\n[2] ...\n\nPhase 7: Deliver\nSave final report to memory/research/[topic]-终极版-[date].md\nGenerate PDF via pymupdf and save to ~/.openclaw/media/outbound/\nSend PDF to user via message tool\nQuality Standards\nMinimum sources: 15 unique URLs per model (60+ total across 4 models)\nSource diversity: No more than 3 citations from same domain per model\nFreshness: Prefer sources < 6 months old; flag older data\nCross-validation: Key claims must appear in 2+ models' findings\nFramework compliance: Every framework component must be addressed\nConfidence scoring: High (3-4 models agree + strong sources), Medium (2 models or weak sources), Low (1 model or no source)\nNo hallucination: Every factual claim must have a source\nAdaptation by Topic Type\nFinancial / Stock Research\nFrameworks: DCF, Comparable Analysis, Power Law\nCheck SEC/regulatory filings, earnings transcripts\nInclude key metrics (revenue, margins, P/E, debt)\nSee references/financial-research.md\nMarket / Industry Research\nFrameworks: Porter's Five Forces, TAM/SAM/SOM, 7 Powers\nCompetitive landscape, key players, market share\nApply Winner-Takes-More thesis where relevant\nStrategy / Business Model\nFrameworks: Schwerpunkt/High Ground, Business Model Canvas, JTBD\nIdentify the constraint, the scarce asset, expansion path\nCompare to historical precedents (Rockefeller, Ramp, etc.)\nTechnical / Product Research\nFrameworks: Wardley Maps, Build vs Buy, Gartner Hype Cycle\nArchitecture, benchmarks, alternatives matrix\nCommunity sentiment (GitHub, HN, Reddit)"
  },
  "trust": {
    "sourceLabel": "tencent",
    "provenanceUrl": "https://clawhub.ai/LeadingOT/opusflame-deep-research",
    "publisherUrl": "https://clawhub.ai/LeadingOT/opusflame-deep-research",
    "owner": "LeadingOT",
    "version": "2.0.0",
    "license": null,
    "verificationStatus": "Indexed source record"
  },
  "links": {
    "detailUrl": "https://openagent3.xyz/skills/opusflame-deep-research",
    "downloadUrl": "https://openagent3.xyz/downloads/opusflame-deep-research",
    "agentUrl": "https://openagent3.xyz/skills/opusflame-deep-research/agent",
    "manifestUrl": "https://openagent3.xyz/skills/opusflame-deep-research/agent.json",
    "briefUrl": "https://openagent3.xyz/skills/opusflame-deep-research/agent.md"
  }
}