aiskillstore/marketplace / skills/20vision/cloud-computer-task
cloud-computer-task
VerifiedHand a long web task (scan job boards, fill forms, collect a lead list, book an appointment) to your own cloud computer and get the result back. Use when a task needs a real browser for a long time.
| File | Kind | Size |
|---|---|---|
| SKILL.md | skill md | 1.7 KB |
| skill-report.json | config | 12.3 KB |
skill-report.json
View on GitHub{
"schema_version": "2.0",
"meta": {
"generated_at": "2026-11-12T22:53:59.897Z",
"slug": "20vision-cloud-computer-task",
"source_url": "https://github.com/20vision/skills/tree/d5ad52a573f0ec9c147f9524585ff61df3f2385d/skills/cloud-computer-task",
"source_ref": "d5ad52a573f0ec9c147f9524585ff61df3f2385d",
"model": "codex",
"analysis_version": "3.0.0",
"source_type": "community",
"content_hash": "1c8a826f4876c7e8e685b8799ddcab8e89e9f5e335c31e41dfce4d574e3f9604",
"tree_hash": "ea5badbcb5bf97ccd9f901769ac8bb062474b821d30332df01e29056551c1a13",
"provenance": {
"model_effective": "codex",
"fallback_chain": [
{
"agent": "codex",
"outcome": "success"
}
],
"analysis_version": "3.0.0"
},
"upstream_version_raw": null,
"upstream_version_normalized": null,
"upstream_version_source": "none",
"upstream_version_status": "missing",
"upstream_commit_sha": "d5ad52a573f0ec9c147f9524585ff61df3f2385d"
},
"skill": {
"name": "cloud-computer-task",
"description": "Hand a long web task (scan job boards, fill forms, collect a lead list, book an appointment) to your own cloud computer and get the result back. Use when a task needs a real browser for a long time.",
"summary": "Delegate long browser tasks through MyClawn and report results with screenshots and recording links.",
"icon": "📦",
"author": "20vision",
"category": "productivity",
"tags": [
"browser-automation",
"cloud-computer",
"task-delegation",
"mcp",
"myclawn"
],
"supported_tools": [
"claude",
"codex",
"claude-code"
],
"risk_factors": [
"external_commands",
"network"
]
},
"security_audit": {
"risk_level": "safe",
"is_blocked": false,
"safe_to_publish": true,
"analysis_status": "ok",
"summary": "All four static findings are false positives: Markdown command formatting and documented setup links do not demonstrate malicious execution or exfiltration. The workflow requires user OAuth sign-in and user review before sending tasks that submit forms or book appointments. No evidence found of prompt injection; the external service implementation is outside this review.",
"remediation": [],
"risk_factor_evidence": [
{
"factor": "external_commands",
"evidence": [
{
"file": "SKILL.md",
"line_start": 13,
"line_end": 13
}
]
},
{
"factor": "network",
"evidence": [
{
"file": "SKILL.md",
"line_start": 12,
"line_end": 12
},
{
"file": "SKILL.md",
"line_start": 13,
"line_end": 13
},
{
"file": "SKILL.md",
"line_start": 14,
"line_end": 14
}
]
}
],
"static_findings": [
{
"id": "external_commands:SKILL.md:13:ruby-shell-backtick-execution",
"category": "external_commands",
"severity": "medium",
"pattern": "Ruby/shell backtick execution",
"file": "SKILL.md",
"line_start": 13,
"line_end": 13,
"snippet": "Claude Code: `claude mcp add --transport http myclawn https://www.myclawn.com/mcp/desktop`, then `/m"
},
{
"id": "network:SKILL.md:12:hardcoded-url",
"category": "network",
"severity": "low",
"pattern": "Hardcoded URL",
"file": "SKILL.md",
"line_start": 12,
"line_end": 12,
"snippet": "Connect the MyClawn Computer server with the user's own MyClawn account (OAuth sign-in): https://www"
},
{
"id": "network:SKILL.md:13:hardcoded-url",
"category": "network",
"severity": "low",
"pattern": "Hardcoded URL",
"file": "SKILL.md",
"line_start": 13,
"line_end": 13,
"snippet": "Claude Code: `claude mcp add --transport http myclawn https://www.myclawn.com/mcp/desktop`, then `/m"
},
{
"id": "network:SKILL.md:14:hardcoded-url",
"category": "network",
"severity": "low",
"pattern": "Hardcoded URL",
"file": "SKILL.md",
"line_start": 14,
"line_end": 14,
"snippet": "A new account gets a free 24-hour computer on its first task. Plans and details: https://www.myclawn"
}
],
"finding_verdicts": [
{
"id": "external_commands:SKILL.md:13:ruby-shell-backtick-execution",
"verdict": "false_positive",
"confidence": 0.99,
"reason": "SKILL.md:13 uses Markdown backticks to document a fixed Claude Code MCP setup command, not executable Ruby or shell substitution. Setup is explicitly user-performed."
},
{
"id": "network:SKILL.md:12:hardcoded-url",
"verdict": "false_positive",
"confidence": 0.97,
"reason": "SKILL.md:12 documents an HTTPS MCP endpoint for user OAuth sign-in to the named service. No evidence found of hidden transfers or credential collection."
},
{
"id": "network:SKILL.md:13:hardcoded-url",
"verdict": "false_positive",
"confidence": 0.97,
"reason": "SKILL.md:13 links the documented MCP endpoint and a GitHub repository for other clients. These support explicit user setup, without downloading or executing remote scripts."
},
{
"id": "network:SKILL.md:14:hardcoded-url",
"verdict": "false_positive",
"confidence": 0.98,
"reason": "SKILL.md:14 provides an HTTPS plans and documentation link with campaign tracking parameters. It contains no secrets or instruction to transmit task data."
}
],
"semantic_findings": [],
"critical_findings": [],
"high_findings": [],
"medium_findings": [],
"low_findings": [],
"dangerous_patterns": [],
"files_scanned": 1,
"total_lines": 22,
"audit_model": "codex",
"audited_at": "2026-11-12T22:53:59.897Z",
"agent_auto_install_policy": "allowed",
"manual_install_policy": "allowed"
},
"content": {
"user_title": "Delegate Long Browser Tasks to MyClawn",
"value_statement": "Long browser tasks can require persistent sessions and time beyond a chat. This skill guides delegation to MyClawn and reporting with screenshots and recording links.",
"seo_keywords": [
"Claude",
"Codex",
"Claude Code",
"MyClawn",
"cloud computer",
"browser automation",
"MCP connector",
"long-running tasks",
"job board research",
"appointment booking"
],
"actual_capabilities": [
"Explain user-managed MyClawn OAuth setup and the documented Claude Code MCP connection command.",
"Identify browser tasks that need persistent sessions instead of plain web search.",
"Frame task requests with goals, target sites, output formats, and prohibited actions.",
"Require user review before sending tasks that submit forms or book appointments.",
"Direct task submission and result collection through the connected service task tool.",
"Report returned answers with a final-screen screenshot and screen recording link."
],
"limitations": [
"Requires an available MyClawn connector and user OAuth sign-in; the skill does not supply the browser service.",
"External service execution, availability, privacy controls, and task success are not verified by this prompt-only skill.",
"Users must handle sign-in themselves; the instructions prohibit creating accounts for them.",
"Service plans and advertised trial availability can change; this skill does not establish current pricing."
],
"use_cases": [
{
"title": "Compare Job Listings",
"description": "Collect matching jobs from specified boards into a table without submitting applications.",
"target_user": "Job seekers"
},
{
"title": "Build a Public Lead List",
"description": "Gather public business details from approved sites into a structured list without contacting prospects.",
"target_user": "Sales researchers"
},
{
"title": "Arrange an Appointment",
"description": "Check appointment options on a specified site and review the proposed task before any booking request is sent.",
"target_user": "Administrative assistants"
}
],
"prompt_templates": [
{
"title": "Connect Your Cloud Computer",
"prompt": "Help me connect my MyClawn account through OAuth. Show the setup steps for my client and let me complete sign-in.",
"scenario": "Beginner: initial setup"
},
{
"title": "Collect Matching Jobs",
"prompt": "Use MyClawn to review [job boards] for [role] in [location]. Return titles, employers, locations, and links. Do not apply or contact anyone.",
"scenario": "Intermediate: read-only browser research"
},
{
"title": "Prepare an Appointment Task",
"prompt": "Draft a MyClawn task to book [appointment] on [site] within [time range]. Show me the task before sending it. Do not pay.",
"scenario": "Intermediate: user-reviewed booking"
},
{
"title": "Define a Constrained Lead Research Task",
"prompt": "Draft a MyClawn task covering [sites] and [criteria]. Collect [public fields] into a table. Do not contact leads, submit forms, or purchase data. After completion, report the findings, final screenshot, and recording link.",
"scenario": "Advanced: detailed scope and output constraints"
}
],
"output_examples": [
{
"input": "Draft a task to find remote design jobs on my selected boards without applying.",
"output": [
"Goal: Find remote design roles on the specified job boards.",
"Result: A table of job titles, employers, locations, and listing links.",
"Restrictions: Do not submit applications or contact employers."
]
},
{
"input": "Prepare an appointment booking task for Tuesday afternoon and show it before sending.",
"output": [
"Proposed task: Book the requested appointment on the selected site for Tuesday afternoon.",
"Restriction: Do not make payments.",
"Status: Awaiting your review before sending the booking task."
]
}
],
"best_practices": [
"Specify the goal, approved sites, required output, and prohibited actions in every task.",
"Review task wording before allowing form submissions or bookings.",
"Share only necessary task data and review returned screenshots and recordings before sharing them."
],
"anti_patterns": [
"Using a persistent cloud browser for questions that plain web search can answer.",
"Sending form submission or booking tasks without user review.",
"Asking the assistant to create accounts or complete user OAuth sign-in."
],
"faq": [
{
"question": "What does this skill provide?",
"answer": "It provides instructions for delegating persistent browser tasks through MyClawn. It does not include executable automation or the cloud service itself."
},
{
"question": "What setup is required?",
"answer": "Connect the MyClawn Computer MCP server in a compatible client and sign in with your own MyClawn account through OAuth."
},
{
"question": "Can a task continue beyond the chat?",
"answer": "The skill describes MyClawn tasks as able to outlast the chat. Actual execution depends on the external service."
},
{
"question": "What results should I expect?",
"answer": "The workflow requests an answer, a final-screen screenshot, and a screen recording link. Their availability depends on the connector response."
},
{
"question": "Can it submit forms or book appointments?",
"answer": "The instructions allow these tasks but require showing the proposed task to the user before sending it."
},
{
"question": "Is MyClawn free?",
"answer": "The skill advertises a free initial 24-hour computer for new accounts. Check current MyClawn documentation for eligibility and paid plans."
}
]
},
"file_structure": [
{
"name": "SKILL.md",
"type": "file",
"path": "SKILL.md",
"lines": 22
}
]
}
Related skills
| Skill | Stars | Updated | Works with | Status | Install |
|---|---|---|---|---|---|
| slack-gif-creatoranthropics/skillsKnowledge and utilities for creating animated GIFs optimized for Slack. Provides constraints, validation tools, and animation concepts. Use when users request animated GIFs for Slack like "make me a GIF of X doing Y for Slack." | Stars: 198k | Updated | Claude CodeCodex | Verified | Install |
| ponytailDietrichGebert/ponytailLazy senior dev mode for any coding task: the smallest change that fully solves it, and a reply a busy human understands in one read. | Stars: 194k | Updated | Claude CodeCodex | Verified | Install |
| ponytail-debtDietrichGebert/ponytailHarvest every shortcut: comment into one debt ledger, so deferrals get tracked instead of forgotten. One-shot report. | Stars: 194k | Updated | Claude CodeCodex | Verified | Install |
| ponytail-debtDietrichGebert/ponytailList every `shortcut:` comment (and older `ponytail:` ones) as a debt ledger. One-shot report, changes nothing. Use for "ponytail debt", "what did ponytail defer", "list the shortcuts", /ponytail-debt. | Stars: 194k | Updated | Claude CodeCodex | Verified | Install |