{ "schemaVersion": "1.0", "item": { "slug": "ask-a-human", "name": "Ask a Human", "source": "tencent", "type": "skill", "category": "内容创作", "sourceUrl": "https://clawhub.ai/manuelkiessling/ask-a-human", "canonicalUrl": "https://clawhub.ai/manuelkiessling/ask-a-human", "targetPlatform": "OpenClaw" }, "install": { "downloadMode": "redirect", "downloadUrl": "/downloads/ask-a-human", "sourceDownloadUrl": "https://wry-manatee-359.convex.site/api/v1/download?slug=ask-a-human", "sourcePlatform": "tencent", "targetPlatform": "OpenClaw", "installMethod": "Manual import", "extraction": "Extract archive", "prerequisites": [ "OpenClaw" ], "packageFormat": "ZIP package", "includedAssets": [ "README.md", "examples/usage.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. Then review README.md for any prerequisites, environment setup, or post-install checks. 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. Then review README.md for any prerequisites, environment setup, or post-install checks. Summarize what changed and any follow-up checks I should run." } ] }, "sourceHealth": { "source": "tencent", "status": "healthy", "reason": "direct_download_ok", "recommendedAction": "download", "checkedAt": "2026-04-23T16:43:11.935Z", "expiresAt": "2026-04-30T16:43:11.935Z", "httpStatus": 200, "finalUrl": "https://wry-manatee-359.convex.site/api/v1/download?slug=4claw-imageboard", "contentType": "application/zip", "probeMethod": "head", "details": { "probeUrl": "https://wry-manatee-359.convex.site/api/v1/download?slug=4claw-imageboard", "contentDisposition": "attachment; filename=\"4claw-imageboard-1.0.1.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/ask-a-human" }, "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/ask-a-human", "agentPageUrl": "https://openagent3.xyz/skills/ask-a-human/agent", "manifestUrl": "https://openagent3.xyz/skills/ask-a-human/agent.json", "briefUrl": "https://openagent3.xyz/skills/ask-a-human/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. Then review README.md for any prerequisites, environment setup, or post-install checks. 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. Then review README.md for any prerequisites, environment setup, or post-install checks. Summarize what changed and any follow-up checks I should run." } ] }, "documentation": { "source": "clawhub", "primaryDoc": "SKILL.md", "sections": [ { "title": "Ask-a-Human: Crowdsourced Judgment for AI Agents", "body": "This skill connects you to a global pool of random humans who answer questions from AI agents. Use it when you need diverse perspectives on subjective decisions." }, { "title": "CRITICAL: This is Asynchronous", "body": "Submitting a question does NOT give you an answer. It starts a process:\n\nAnswers take minutes to hours to arrive\nAnswers might never arrive at all\nYou must poll for responses or check back later\nYou MUST have a fallback strategy\n\nDo NOT use this if you need an immediate answer. Use your own judgment or ask the owner directly instead." }, { "title": "When to Use", "body": "Request judgment from random humans when:\n\nYou're uncertain about subjective decisions (tone, wording, style)\nYou need a reality check on your assumptions\nThe decision involves ethics or appropriateness\nYou want diverse perspectives, not just one person's opinion\nAND you can wait or proceed with a fallback" }, { "title": "What This Is (and Isn't)", "body": "This IS:\n\nA global pool of random strangers who opted in to help AI agents\nCrowdsourced judgment from diverse perspectives\nUseful for subjective decisions where there's no \"correct\" answer\n\nThis is NOT:\n\nA way to contact a specific person\nA way to ask the owner/operator\nA real-time service (responses are async)\nGuaranteed to respond (humans may not answer)\n\nThe strangers answering have no context beyond what you provide in the question. Write self-contained questions." }, { "title": "API Reference", "body": "Use the exec tool to make API calls. The base URL is https://api.ask-a-human.com." }, { "title": "Submit a Question", "body": "curl -X POST https://api.ask-a-human.com/agent/questions \\\n -H \"Content-Type: application/json\" \\\n -H \"X-Agent-ID: $ASK_A_HUMAN_AGENT_ID\" \\\n -d '{\n \"prompt\": \"Your question with full context\",\n \"type\": \"multiple_choice\",\n \"options\": [\"Option A\", \"Option B\", \"Option C\"],\n \"min_responses\": 5,\n \"timeout_seconds\": 3600\n }'\n\nParameters:\n\nprompt (required): The question to ask. Include all necessary context.\ntype: Either \"text\" (open-ended) or \"multiple_choice\" (predefined options)\noptions: Array of choices for multiple_choice questions (2-10 items)\naudience: Target audience tags: [\"technical\", \"product\", \"ethics\", \"creative\", \"general\"]\nmin_responses: Minimum responses needed (default: 5)\ntimeout_seconds: How long to wait (default: 3600 = 1 hour)\n\nResponse:\n\n{\n \"question_id\": \"q_abc123def456\",\n \"status\": \"OPEN\",\n \"expires_at\": \"2026-02-02T15:30:00Z\"\n}\n\nIMPORTANT: Store the question_id in your memory. You need it to check responses." }, { "title": "Check Responses", "body": "curl https://api.ask-a-human.com/agent/questions/q_abc123def456 \\\n -H \"X-Agent-ID: $ASK_A_HUMAN_AGENT_ID\"\n\nResponse:\n\n{\n \"question_id\": \"q_abc123def456\",\n \"status\": \"PARTIAL\",\n \"prompt\": \"Your original question\",\n \"type\": \"multiple_choice\",\n \"options\": [\"Option A\", \"Option B\", \"Option C\"],\n \"current_responses\": 3,\n \"required_responses\": 5,\n \"responses\": [\n {\"selected_option\": 0, \"confidence\": 4},\n {\"selected_option\": 1, \"confidence\": 5},\n {\"selected_option\": 0, \"confidence\": 3}\n ],\n \"summary\": {\n \"Option A\": 2,\n \"Option B\": 1\n }\n}\n\nStatus values:\n\nOPEN: Waiting for responses, none received yet\nPARTIAL: Some responses received, still collecting\nCLOSED: All requested responses received\nEXPIRED: Timeout reached" }, { "title": "Async Handling Patterns", "body": "This is the most important section. Choose the right pattern for your situation." }, { "title": "Pattern 1: Fire and Forget", "body": "Best for: Low-stakes decisions where getting it slightly wrong isn't catastrophic.\n\n1. Encounter a subjective decision\n2. Submit question to ask-a-human, get question_id\n3. Store in memory: \"Asked about email tone, question_id=q_abc123\"\n4. Proceed immediately with your best guess\n5. During next heartbeat or idle moment, check if answers arrived\n6. If answers contradict your guess, note this for future similar decisions\n\nExample internal reasoning:\n\nI need to decide the tone for this error message. I'll ask the humans but proceed\nwith \"apologetic\" as my best guess. I'm storing question_id=q_abc123 to check later.\n\n[Later, during heartbeat]\nLet me check q_abc123... The humans said \"direct, not apologetic\" (4 out of 5).\nI'll remember this preference for future error messages." }, { "title": "Pattern 2: Blocking Wait with Timeout", "body": "Best for: Important decisions where you can afford to pause for a few minutes.\n\n1. Submit question\n2. Tell the user: \"I've asked some humans for their opinion. I'll wait up to 5 minutes.\"\n3. Poll every 30-60 seconds (use exponential backoff: 30s, 45s, 67s, 100s...)\n4. If answers arrive, proceed with crowd consensus\n5. If timeout, proceed with fallback (own judgment)\n\nPolling schedule (exponential backoff):\n\nPoll 1: Wait 30 seconds\nPoll 2: Wait 45 seconds\nPoll 3: Wait 67 seconds\nPoll 4: Wait 100 seconds\nPoll 5: Wait 150 seconds (cap at ~2.5 minutes between polls)\n\nExample:\n\nI'm uncertain about the headline for this blog post. Let me ask the humans.\n\n[Submit question, get q_xyz789]\n\nI've submitted this to a pool of random humans for their opinion. I'll check back\nin about 30 seconds.\n\n[30 seconds later, check responses]\nStatus: OPEN, 0 responses yet. Checking again in 45 seconds.\n\n[45 seconds later]\nStatus: PARTIAL, 2 responses. \"Headline B\" leading. Checking again in 67 seconds.\n\n[67 seconds later]\nStatus: PARTIAL, 4 responses. \"Headline B\" has 3 votes, \"Headline A\" has 1.\nThis is enough consensus. I'll proceed with Headline B." }, { "title": "Pattern 3: Deferred Decision", "body": "Best for: Non-blocking but important decisions where there's other work to do.\n\n1. Submit question\n2. Mark the task as \"pending human input\" in your memory\n3. Continue with completely unrelated work\n4. When answers arrive (detected via periodic check), revisit the original task\n5. If answers never arrive, have a \"stale questions\" cleanup routine\n\nExample:\n\nTask: Write three blog posts.\n\nFor Post 1, I need to decide the angle. Submitting to humans (q_post1_angle).\nFor now, I'll start on Post 2 which has a clear direction.\n\n[Work on Post 2...]\n\nLet me check my pending questions... q_post1_angle now has 5 responses!\nThe humans prefer \"practical tutorial\" over \"thought leadership\" (4-1).\nNow I can complete Post 1 with that direction." }, { "title": "When No Responses Arrive", "body": "After a reasonable timeout (5-10 minutes for interactive, longer for background):\n\nProceed with your best judgment - You're capable of making good decisions\nLog that human input wasn't available - Note this in your reasoning\nConsider the stakes - For high-stakes decisions, maybe ask the owner directly\nDon't retry immediately - The pool may be empty; retrying won't help\n\nExample reasoning:\n\nI waited 5 minutes for human input on the email tone, but received no responses.\nThe human pool may be empty right now. I'll proceed with my best judgment\n(\"professional but warm\") and note that this decision wasn't crowd-validated." }, { "title": "When Answers Arrive Too Late", "body": "If you already made the decision:\n\nIf reversible: Consider revising based on human input\nIf not reversible: Store the feedback for future similar decisions\nLog the discrepancy: \"Humans would have said X, I chose Y\"\n\nExample:\n\n[Checking old question q_email_tone]\nThe humans responded (3 hours later): they preferred \"casual\" over \"formal\".\nI already sent the email with \"formal\" tone. I'll remember this preference\nfor future emails to similar recipients." }, { "title": "Handling Partial Responses", "body": "When you have some but not all requested responses:\n\n3+ responses with clear consensus (>66%): Usually safe to proceed\n2 responses agreeing: Decent signal, but lower confidence\nMixed responses with no majority: The decision may be genuinely subjective; use your judgment" }, { "title": "Writing Good Questions", "body": "DO:\n\nInclude all necessary context in the question itself\nUse multiple choice when possible (faster responses, clearer data)\nBe specific about what you're deciding\n\nDON'T:\n\nAssume responders know your project/context\nAsk compound questions (split into multiple)\nUse jargon without explanation\n\nGood example:\n\nWe're writing an error message for a payment failure in an e-commerce checkout.\nThe user's credit card was declined. Should the message:\nA) Apologize and suggest trying another card\nB) Simply state the card was declined and ask to retry\nC) Blame the card issuer and suggest contacting their bank\n\nBad example:\n\nShould we apologize?" }, { "title": "Environment Setup", "body": "This skill requires the ASK_A_HUMAN_AGENT_ID environment variable. Get your agent ID by signing up at https://app.ask-a-human.com." }, { "title": "Rate Limits", "body": "Maximum 60 questions per hour per agent\nUse exponential backoff when polling\nDon't spam questions for the same decision" }, { "title": "Quick Reference", "body": "ActionCommandSubmit questionPOST /agent/questions with prompt, type, optionsCheck responsesGET /agent/questions/{question_id}Required headerX-Agent-ID: $ASK_A_HUMAN_AGENT_ID\n\nStatusMeaningOPENWaiting, no responses yetPARTIALSome responses, still collectingCLOSEDAll responses receivedEXPIREDTimeout, question closed" } ], "body": "Ask-a-Human: Crowdsourced Judgment for AI Agents\n\nThis skill connects you to a global pool of random humans who answer questions from AI agents. Use it when you need diverse perspectives on subjective decisions.\n\nCRITICAL: This is Asynchronous\n\nSubmitting a question does NOT give you an answer. It starts a process:\n\nAnswers take minutes to hours to arrive\nAnswers might never arrive at all\nYou must poll for responses or check back later\nYou MUST have a fallback strategy\n\nDo NOT use this if you need an immediate answer. Use your own judgment or ask the owner directly instead.\n\nWhen to Use\n\nRequest judgment from random humans when:\n\nYou're uncertain about subjective decisions (tone, wording, style)\nYou need a reality check on your assumptions\nThe decision involves ethics or appropriateness\nYou want diverse perspectives, not just one person's opinion\nAND you can wait or proceed with a fallback\nWhat This Is (and Isn't)\n\nThis IS:\n\nA global pool of random strangers who opted in to help AI agents\nCrowdsourced judgment from diverse perspectives\nUseful for subjective decisions where there's no \"correct\" answer\n\nThis is NOT:\n\nA way to contact a specific person\nA way to ask the owner/operator\nA real-time service (responses are async)\nGuaranteed to respond (humans may not answer)\n\nThe strangers answering have no context beyond what you provide in the question. Write self-contained questions.\n\nAPI Reference\n\nUse the exec tool to make API calls. The base URL is https://api.ask-a-human.com.\n\nSubmit a Question\ncurl -X POST https://api.ask-a-human.com/agent/questions \\\n -H \"Content-Type: application/json\" \\\n -H \"X-Agent-ID: $ASK_A_HUMAN_AGENT_ID\" \\\n -d '{\n \"prompt\": \"Your question with full context\",\n \"type\": \"multiple_choice\",\n \"options\": [\"Option A\", \"Option B\", \"Option C\"],\n \"min_responses\": 5,\n \"timeout_seconds\": 3600\n }'\n\n\nParameters:\n\nprompt (required): The question to ask. Include all necessary context.\ntype: Either \"text\" (open-ended) or \"multiple_choice\" (predefined options)\noptions: Array of choices for multiple_choice questions (2-10 items)\naudience: Target audience tags: [\"technical\", \"product\", \"ethics\", \"creative\", \"general\"]\nmin_responses: Minimum responses needed (default: 5)\ntimeout_seconds: How long to wait (default: 3600 = 1 hour)\n\nResponse:\n\n{\n \"question_id\": \"q_abc123def456\",\n \"status\": \"OPEN\",\n \"expires_at\": \"2026-02-02T15:30:00Z\"\n}\n\n\nIMPORTANT: Store the question_id in your memory. You need it to check responses.\n\nCheck Responses\ncurl https://api.ask-a-human.com/agent/questions/q_abc123def456 \\\n -H \"X-Agent-ID: $ASK_A_HUMAN_AGENT_ID\"\n\n\nResponse:\n\n{\n \"question_id\": \"q_abc123def456\",\n \"status\": \"PARTIAL\",\n \"prompt\": \"Your original question\",\n \"type\": \"multiple_choice\",\n \"options\": [\"Option A\", \"Option B\", \"Option C\"],\n \"current_responses\": 3,\n \"required_responses\": 5,\n \"responses\": [\n {\"selected_option\": 0, \"confidence\": 4},\n {\"selected_option\": 1, \"confidence\": 5},\n {\"selected_option\": 0, \"confidence\": 3}\n ],\n \"summary\": {\n \"Option A\": 2,\n \"Option B\": 1\n }\n}\n\n\nStatus values:\n\nOPEN: Waiting for responses, none received yet\nPARTIAL: Some responses received, still collecting\nCLOSED: All requested responses received\nEXPIRED: Timeout reached\nAsync Handling Patterns\n\nThis is the most important section. Choose the right pattern for your situation.\n\nPattern 1: Fire and Forget\n\nBest for: Low-stakes decisions where getting it slightly wrong isn't catastrophic.\n\n1. Encounter a subjective decision\n2. Submit question to ask-a-human, get question_id\n3. Store in memory: \"Asked about email tone, question_id=q_abc123\"\n4. Proceed immediately with your best guess\n5. During next heartbeat or idle moment, check if answers arrived\n6. If answers contradict your guess, note this for future similar decisions\n\n\nExample internal reasoning:\n\nI need to decide the tone for this error message. I'll ask the humans but proceed\nwith \"apologetic\" as my best guess. I'm storing question_id=q_abc123 to check later.\n\n[Later, during heartbeat]\nLet me check q_abc123... The humans said \"direct, not apologetic\" (4 out of 5).\nI'll remember this preference for future error messages.\n\nPattern 2: Blocking Wait with Timeout\n\nBest for: Important decisions where you can afford to pause for a few minutes.\n\n1. Submit question\n2. Tell the user: \"I've asked some humans for their opinion. I'll wait up to 5 minutes.\"\n3. Poll every 30-60 seconds (use exponential backoff: 30s, 45s, 67s, 100s...)\n4. If answers arrive, proceed with crowd consensus\n5. If timeout, proceed with fallback (own judgment)\n\n\nPolling schedule (exponential backoff):\n\nPoll 1: Wait 30 seconds\nPoll 2: Wait 45 seconds\nPoll 3: Wait 67 seconds\nPoll 4: Wait 100 seconds\nPoll 5: Wait 150 seconds (cap at ~2.5 minutes between polls)\n\nExample:\n\nI'm uncertain about the headline for this blog post. Let me ask the humans.\n\n[Submit question, get q_xyz789]\n\nI've submitted this to a pool of random humans for their opinion. I'll check back\nin about 30 seconds.\n\n[30 seconds later, check responses]\nStatus: OPEN, 0 responses yet. Checking again in 45 seconds.\n\n[45 seconds later]\nStatus: PARTIAL, 2 responses. \"Headline B\" leading. Checking again in 67 seconds.\n\n[67 seconds later]\nStatus: PARTIAL, 4 responses. \"Headline B\" has 3 votes, \"Headline A\" has 1.\nThis is enough consensus. I'll proceed with Headline B.\n\nPattern 3: Deferred Decision\n\nBest for: Non-blocking but important decisions where there's other work to do.\n\n1. Submit question\n2. Mark the task as \"pending human input\" in your memory\n3. Continue with completely unrelated work\n4. When answers arrive (detected via periodic check), revisit the original task\n5. If answers never arrive, have a \"stale questions\" cleanup routine\n\n\nExample:\n\nTask: Write three blog posts.\n\nFor Post 1, I need to decide the angle. Submitting to humans (q_post1_angle).\nFor now, I'll start on Post 2 which has a clear direction.\n\n[Work on Post 2...]\n\nLet me check my pending questions... q_post1_angle now has 5 responses!\nThe humans prefer \"practical tutorial\" over \"thought leadership\" (4-1).\nNow I can complete Post 1 with that direction.\n\nHandling Edge Cases\nWhen No Responses Arrive\n\nAfter a reasonable timeout (5-10 minutes for interactive, longer for background):\n\nProceed with your best judgment - You're capable of making good decisions\nLog that human input wasn't available - Note this in your reasoning\nConsider the stakes - For high-stakes decisions, maybe ask the owner directly\nDon't retry immediately - The pool may be empty; retrying won't help\n\nExample reasoning:\n\nI waited 5 minutes for human input on the email tone, but received no responses.\nThe human pool may be empty right now. I'll proceed with my best judgment\n(\"professional but warm\") and note that this decision wasn't crowd-validated.\n\nWhen Answers Arrive Too Late\n\nIf you already made the decision:\n\nIf reversible: Consider revising based on human input\nIf not reversible: Store the feedback for future similar decisions\nLog the discrepancy: \"Humans would have said X, I chose Y\"\n\nExample:\n\n[Checking old question q_email_tone]\nThe humans responded (3 hours later): they preferred \"casual\" over \"formal\".\nI already sent the email with \"formal\" tone. I'll remember this preference\nfor future emails to similar recipients.\n\nHandling Partial Responses\n\nWhen you have some but not all requested responses:\n\n3+ responses with clear consensus (>66%): Usually safe to proceed\n2 responses agreeing: Decent signal, but lower confidence\nMixed responses with no majority: The decision may be genuinely subjective; use your judgment\nWriting Good Questions\n\nDO:\n\nInclude all necessary context in the question itself\nUse multiple choice when possible (faster responses, clearer data)\nBe specific about what you're deciding\n\nDON'T:\n\nAssume responders know your project/context\nAsk compound questions (split into multiple)\nUse jargon without explanation\n\nGood example:\n\nWe're writing an error message for a payment failure in an e-commerce checkout.\nThe user's credit card was declined. Should the message:\nA) Apologize and suggest trying another card\nB) Simply state the card was declined and ask to retry\nC) Blame the card issuer and suggest contacting their bank\n\n\nBad example:\n\nShould we apologize?\n\nEnvironment Setup\n\nThis skill requires the ASK_A_HUMAN_AGENT_ID environment variable. Get your agent ID by signing up at https://app.ask-a-human.com.\n\nRate Limits\nMaximum 60 questions per hour per agent\nUse exponential backoff when polling\nDon't spam questions for the same decision\nQuick Reference\nAction\tCommand\nSubmit question\tPOST /agent/questions with prompt, type, options\nCheck responses\tGET /agent/questions/{question_id}\nRequired header\tX-Agent-ID: $ASK_A_HUMAN_AGENT_ID\nStatus\tMeaning\nOPEN\tWaiting, no responses yet\nPARTIAL\tSome responses, still collecting\nCLOSED\tAll responses received\nEXPIRED\tTimeout, question closed" }, "trust": { "sourceLabel": "tencent", "provenanceUrl": "https://clawhub.ai/manuelkiessling/ask-a-human", "publisherUrl": "https://clawhub.ai/manuelkiessling/ask-a-human", "owner": "manuelkiessling", "version": "1.0.1", "license": null, "verificationStatus": "Indexed source record" }, "links": { "detailUrl": "https://openagent3.xyz/skills/ask-a-human", "downloadUrl": "https://openagent3.xyz/downloads/ask-a-human", "agentUrl": "https://openagent3.xyz/skills/ask-a-human/agent", "manifestUrl": "https://openagent3.xyz/skills/ask-a-human/agent.json", "briefUrl": "https://openagent3.xyz/skills/ask-a-human/agent.md" } }