{ "schemaVersion": "1.0", "item": { "slug": "research-agent", "name": "Research", "source": "tencent", "type": "skill", "category": "开发工具", "sourceUrl": "https://clawhub.ai/BrennerSpear/research-agent", "canonicalUrl": "https://clawhub.ai/BrennerSpear/research-agent", "targetPlatform": "OpenClaw" }, "install": { "downloadMode": "redirect", "downloadUrl": "/downloads/research-agent", "sourceDownloadUrl": "https://wry-manatee-359.convex.site/api/v1/download?slug=research-agent", "sourcePlatform": "tencent", "targetPlatform": "OpenClaw", "installMethod": "Manual import", "extraction": "Extract archive", "prerequisites": [ "OpenClaw" ], "packageFormat": "ZIP package", "includedAssets": [ "OPENCLAW.md", "SETUP.md", "SKILL.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": "research-agent", "status": "healthy", "reason": "direct_download_ok", "recommendedAction": "download", "checkedAt": "2026-05-08T04:55:42.083Z", "expiresAt": "2026-05-15T04:55:42.083Z", "httpStatus": 200, "finalUrl": "https://wry-manatee-359.convex.site/api/v1/download?slug=research-agent", "contentType": "application/zip", "probeMethod": "head", "details": { "probeUrl": "https://wry-manatee-359.convex.site/api/v1/download?slug=research-agent", "contentDisposition": "attachment; filename=\"research-agent-1.0.0.zip\"", "redirectLocation": null, "bodySnippet": null, "slug": "research-agent" }, "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/research-agent" }, "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/research-agent", "agentPageUrl": "https://openagent3.xyz/skills/research-agent/agent", "manifestUrl": "https://openagent3.xyz/skills/research-agent/agent.json", "briefUrl": "https://openagent3.xyz/skills/research-agent/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": "Description", "body": "Conduct open-ended research on a topic, building a living markdown document. The conversation is ephemeral; the document is what matters." }, { "title": "Trigger", "body": "Activate when the user wants to:\n\nResearch a topic, idea, or question\nExplore something before committing to building it\nInvestigate options, patterns, or approaches\nCreate a \"research doc\" or \"investigation\"\nRun deep async research on a complex topic" }, { "title": "Research Directory", "body": "Each research topic gets its own folder:\n\n~/.openclaw/workspace/research//\n├── prompt.md # Original research question/prompt\n├── research.md # Main findings (Parallel output or interactive notes)\n├── research.pdf # PDF export (when generated)\n└── ... # Any other related files (data, images, etc.)" }, { "title": "1. Interactive Research (default)", "body": "For topics you explore together in conversation. You search, synthesize, and update the doc in real-time." }, { "title": "2. Deep Research (async)", "body": "For complex topics that need comprehensive investigation. Uses the Parallel AI API via parallel-research CLI. Takes minutes to hours, returns detailed markdown reports.\n\nWhen to use deep research:\n\nMarket analysis, competitive landscape\nTechnical deep-dives requiring extensive source gathering\nMulti-faceted questions that benefit from parallel exploration\nWhen user says \"deep research\" or wants comprehensive coverage" }, { "title": "1. Initialize Research", "body": "Create the research folder at ~/.openclaw/workspace/research//\n\n\nCreate prompt.md with the original question:\n# \n\n> \n\n**Started:** \n\n\n\nCreate research.md with the working structure:\n# \n\n**Status:** Active Research\n**Started:** \n**Last Updated:** \n\n---\n\n## Open Questions\n- \n\n## Findings\n\n\n## Options / Approaches\n\n\n## Resources\n\n\n## Next Steps\n\n\n\n\nConfirm with user - Show the folder was created and ask what to explore first." }, { "title": "2. Research Loop", "body": "For each exchange:\n\nDo the research - Web search, fetch docs, explore code\nUpdate the document - Add findings, move answered questions, add sources\nShow progress - Note what was added (don't repeat everything)\nPrompt next direction - End with a question or suggestion\n\nKey behaviors:\n\nUpdate existing sections over creating new ones\nUse bullet points for findings; prose for summaries\nNote uncertainty (\"seems like\", \"according to X\", \"unverified\")\nLink to sources whenever possible" }, { "title": "3. Synthesis Checkpoints", "body": "Every 5-10 exchanges, offer to:\n\nWrite a \"Current Understanding\" summary\nPrune redundant findings\nReorganize if unwieldy\nCheck blind spots" }, { "title": "4. Completion", "body": "When research is complete, update the status in research.md:\n\n\"Status: Complete\" — Done, stays in place as reference\n\"Status: Ongoing\" — Living doc, will be updated over time\n\nIf the research is specifically for building a project:\n\nGraduate to ~/specs/.md as a project spec\nOr create a project directly based on findings\nUpdate status to \"Status: Graduated → ~/specs/...\"\n\nMost research is just research — it doesn't need to become a spec. Only graduate if you're actually building something from it." }, { "title": "1. Start Deep Research", "body": "parallel-research create \"Your research question\" --processor ultra --wait\n\nProcessor options:\n\nlite, base, core, pro, ultra (default), ultra2x, ultra4x, ultra8x\nAdd -fast suffix for speed over depth: ultra-fast, pro-fast, etc.\n\nOptions:\n\n-w, --wait — Wait for completion and show result\n-p, --processor — Choose processor tier\n-j, --json — Raw JSON output" }, { "title": "2. Schedule Auto-Check (optional)", "body": "Deep research tasks take minutes to hours. You'll want to poll for results automatically rather than checking manually.\n\nOptions:\n\nOpenClaw users: See OPENCLAW.md for cron-based auto-check scheduling\nOther setups: Use any scheduler (cron, systemd timer, CI job) to periodically run parallel-research status and parallel-research result until complete\nSimple approach: Just use parallel-research create \"...\" --wait to block until done (works for shorter tasks)" }, { "title": "3. Manual Check (if needed)", "body": "parallel-research status \nparallel-research result " }, { "title": "4. Save to Research Folder", "body": "Create the research folder and save results:\n\n~/.openclaw/workspace/research//\n├── prompt.md # Original question + run metadata\n├── research.md # Full Parallel output\n\nprompt.md should include:\n\n# \n\n> \n\n**Run ID:** \n**Processor:** \n**Started:** \n**Completed:** \n\nresearch.md contains the full Parallel output, plus any follow-up notes." }, { "title": "PDF Export", "body": "All PDFs go in the research folder — never save to tmp/. Whether using export-pdf, the browser pdf action, or any other method, the output path must be research//.\n\nUse the export-pdf script to convert research docs to PDF:\n\nexport-pdf ~/.openclaw/workspace/research//research.md\n# Creates: ~/.openclaw/workspace/research//research.pdf\n\nFor browser-generated PDFs (e.g. saving a webpage as PDF):\n\nbrowser pdf → save to research//.pdf\n\nNote: Tables render as stacked rows (PyMuPDF limitation). Acceptable for research docs." }, { "title": "Commands", "body": "\"new research: \" - Start interactive research doc\n\"deep research: \" - Start async deep research\n\"show doc\" / \"show research\" - Display current research file\n\"summarize\" - Synthesis checkpoint\n\"graduate\" - Move research to next phase\n\"archive\" - Mark as complete reference\n\"export pdf\" - Export to PDF\n\"check research\" - Check status of pending deep research tasks" }, { "title": "Document Principles", "body": "Atomic findings - One insight per bullet\nLink everything - Sources, docs, repos\nCapture context - Why did we look at this?\nNote confidence - Use qualifiers when uncertain\nDate important findings - Especially for fast-moving topics" }, { "title": "Setup", "body": "See SETUP.md for first-time installation of:\n\nparallel-research CLI\nPDF export tools (pandoc, PyMuPDF)" } ], "body": "Research Skill\nDescription\n\nConduct open-ended research on a topic, building a living markdown document. The conversation is ephemeral; the document is what matters.\n\nTrigger\n\nActivate when the user wants to:\n\nResearch a topic, idea, or question\nExplore something before committing to building it\nInvestigate options, patterns, or approaches\nCreate a \"research doc\" or \"investigation\"\nRun deep async research on a complex topic\nResearch Directory\n\nEach research topic gets its own folder:\n\n~/.openclaw/workspace/research//\n├── prompt.md # Original research question/prompt\n├── research.md # Main findings (Parallel output or interactive notes)\n├── research.pdf # PDF export (when generated)\n└── ... # Any other related files (data, images, etc.)\n\nTwo Research Modes\n1. Interactive Research (default)\n\nFor topics you explore together in conversation. You search, synthesize, and update the doc in real-time.\n\n2. Deep Research (async)\n\nFor complex topics that need comprehensive investigation. Uses the Parallel AI API via parallel-research CLI. Takes minutes to hours, returns detailed markdown reports.\n\nWhen to use deep research:\n\nMarket analysis, competitive landscape\nTechnical deep-dives requiring extensive source gathering\nMulti-faceted questions that benefit from parallel exploration\nWhen user says \"deep research\" or wants comprehensive coverage\nInteractive Research Workflow\n1. Initialize Research\n\nCreate the research folder at ~/.openclaw/workspace/research//\n\nCreate prompt.md with the original question:\n\n# \n\n> \n\n**Started:** \n\n\nCreate research.md with the working structure:\n\n# \n\n**Status:** Active Research\n**Started:** \n**Last Updated:** \n\n---\n\n## Open Questions\n- \n\n## Findings\n\n\n## Options / Approaches\n\n\n## Resources\n\n\n## Next Steps\n\n\n\nConfirm with user - Show the folder was created and ask what to explore first.\n\n2. Research Loop\n\nFor each exchange:\n\nDo the research - Web search, fetch docs, explore code\nUpdate the document - Add findings, move answered questions, add sources\nShow progress - Note what was added (don't repeat everything)\nPrompt next direction - End with a question or suggestion\n\nKey behaviors:\n\nUpdate existing sections over creating new ones\nUse bullet points for findings; prose for summaries\nNote uncertainty (\"seems like\", \"according to X\", \"unverified\")\nLink to sources whenever possible\n3. Synthesis Checkpoints\n\nEvery 5-10 exchanges, offer to:\n\nWrite a \"Current Understanding\" summary\nPrune redundant findings\nReorganize if unwieldy\nCheck blind spots\n4. Completion\n\nWhen research is complete, update the status in research.md:\n\n\"Status: Complete\" — Done, stays in place as reference\n\"Status: Ongoing\" — Living doc, will be updated over time\n\nIf the research is specifically for building a project:\n\nGraduate to ~/specs/.md as a project spec\nOr create a project directly based on findings\nUpdate status to \"Status: Graduated → ~/specs/...\"\n\nMost research is just research — it doesn't need to become a spec. Only graduate if you're actually building something from it.\n\nDeep Research Workflow\n1. Start Deep Research\nparallel-research create \"Your research question\" --processor ultra --wait\n\n\nProcessor options:\n\nlite, base, core, pro, ultra (default), ultra2x, ultra4x, ultra8x\nAdd -fast suffix for speed over depth: ultra-fast, pro-fast, etc.\n\nOptions:\n\n-w, --wait — Wait for completion and show result\n-p, --processor — Choose processor tier\n-j, --json — Raw JSON output\n2. Schedule Auto-Check (optional)\n\nDeep research tasks take minutes to hours. You'll want to poll for results automatically rather than checking manually.\n\nOptions:\n\nOpenClaw users: See OPENCLAW.md for cron-based auto-check scheduling\nOther setups: Use any scheduler (cron, systemd timer, CI job) to periodically run parallel-research status and parallel-research result until complete\nSimple approach: Just use parallel-research create \"...\" --wait to block until done (works for shorter tasks)\n3. Manual Check (if needed)\nparallel-research status \nparallel-research result \n\n4. Save to Research Folder\n\nCreate the research folder and save results:\n\n~/.openclaw/workspace/research//\n├── prompt.md # Original question + run metadata\n├── research.md # Full Parallel output\n\n\nprompt.md should include:\n\n# \n\n> \n\n**Run ID:** \n**Processor:** \n**Started:** \n**Completed:** \n\n\nresearch.md contains the full Parallel output, plus any follow-up notes.\n\nPDF Export\n\nAll PDFs go in the research folder — never save to tmp/. Whether using export-pdf, the browser pdf action, or any other method, the output path must be research//.\n\nUse the export-pdf script to convert research docs to PDF:\n\nexport-pdf ~/.openclaw/workspace/research//research.md\n# Creates: ~/.openclaw/workspace/research//research.pdf\n\n\nFor browser-generated PDFs (e.g. saving a webpage as PDF):\n\nbrowser pdf → save to research//.pdf\n\n\nNote: Tables render as stacked rows (PyMuPDF limitation). Acceptable for research docs.\n\nCommands\n\"new research: \" - Start interactive research doc\n\"deep research: \" - Start async deep research\n\"show doc\" / \"show research\" - Display current research file\n\"summarize\" - Synthesis checkpoint\n\"graduate\" - Move research to next phase\n\"archive\" - Mark as complete reference\n\"export pdf\" - Export to PDF\n\"check research\" - Check status of pending deep research tasks\nDocument Principles\nAtomic findings - One insight per bullet\nLink everything - Sources, docs, repos\nCapture context - Why did we look at this?\nNote confidence - Use qualifiers when uncertain\nDate important findings - Especially for fast-moving topics\nSetup\n\nSee SETUP.md for first-time installation of:\n\nparallel-research CLI\nPDF export tools (pandoc, PyMuPDF)" }, "trust": { "sourceLabel": "tencent", "provenanceUrl": "https://clawhub.ai/BrennerSpear/research-agent", "publisherUrl": "https://clawhub.ai/BrennerSpear/research-agent", "owner": "BrennerSpear", "version": "1.0.0", "license": null, "verificationStatus": "Indexed source record" }, "links": { "detailUrl": "https://openagent3.xyz/skills/research-agent", "downloadUrl": "https://openagent3.xyz/downloads/research-agent", "agentUrl": "https://openagent3.xyz/skills/research-agent/agent", "manifestUrl": "https://openagent3.xyz/skills/research-agent/agent.json", "briefUrl": "https://openagent3.xyz/skills/research-agent/agent.md" } }