What exactly is broken in popular skills: references to files that are not bundled, actions taken without human approval, processes with no stopping condition. Everything here is measured from the skill text rather than judged by a model, so every line can be checked.
Procedurewill not run
小红书全能助手 — 文案生成、封面制作、内容发布与管理。当用户要求写小红书笔记、生成小红书文案/标题/封面、发小红书、搜索小红书、评论点赞收藏等任何小红书相关操作时使用。支持一站式从文案创作到自动发布的完整流程。封面图使用纯ImageMagick生成,无需API配置。
- References files that are not bundled: {baseDir}/references/title-guide.md, {baseDir}/references/content-guide.md
- 2 referenced file(s) missing: {baseDir}/references/title-guide.md, {baseDir}/references/content-guide.md
- Does not say what the result is
clawhub · ↓ 801
Analyzerwill not run
Run a structured health check of your ecommerce seller account — covering listing quality, policy compliance, performance metrics, and account standing — to catch problems before t
- References files that are not bundled: references/output-template.md, references/platform-health-metrics.md, references/violation-response-guide.md
- 4 referenced file(s) missing: references/output-template.md, references/platform-health-metrics.md, references/violation-response-guide.md
- Does not say what the result is
clawhub · ↓ 851
Generatorwill not run
Grant proposal writing assistant for NIH (R01/R21), NSF and other mainstream funding applications. Triggers when user needs help writing specific aims, research strategy, budget ju
- References files that are not bundled: references/budget_templates.xlsx
- Safety rules and hard prohibitions inside a skill: they belong in the system prompt, here they protect nothing
- 13 top-level sections: this looks like several domains in one skill
clawhub · ↓ 802
Generatorwill not run
Render BestYou health data as visual Dark Glass dashboards via OpenClaw canvas. Daily briefings, action plans, progress snapshots, weekly summaries, meal analysis, and workout plan
- References files that are not bundled: assets/shared.css
- 1 referenced file(s) missing: assets/shared.css
- Does not say what the result is
clawhub · ↓ 801
Analyzerwill not run
嘉为蓝鲸 ITSM 工单分析技能。支持多流程工单数据读取、字段智能映射、新工单处理建议、趋势分析、高频问题识别、SLA 监控。Use when: 需要分析 ITSM 工单数据、为新工单提供处理建议、生成工单日报/周报、识别高频问题、监控 SLA 超时风险。支持不同流程的工单(字段自动适配)。
- References files that are not bundled: references/config.json, references/blueking-api.md, references/ticket-classification.md
- 5 referenced file(s) missing: references/config.json, references/blueking-api.md, references/ticket-classification.md
- Does not say what the result is
clawhub · ↓ 779
Integrationwill not run
Publish posts to Moltbook social network for AI agents. Use when the user wants to create, schedule, or manage posts on Moltbook, handle cross-timezone publishing strategies, manag
- References files that are not bundled: references/moltbook_api.md, references/content_examples.md, references/error_codes.md
- 16 top-level sections: this looks like several domains in one skill
- 5 referenced file(s) missing: references/moltbook_api.md, references/content_examples.md, references/error_codes.md
clawhub · ↓ 802
Generatorwill not run
Model Context Protocol (MCP) bridge for Outline (getoutline.com). Enables AI agents to search, read, create, and manage documents, collections, and comments in an Outline workspace
- References files that are not bundled: url
- 1 referenced file(s) missing: url
- Does not say what the result is
clawhub · ↓ 826
Generatorwill not run
生成初中班会课PPT或英语课PPT。触发条件:(1)用户要求做班会课PPT、主题班会、班主任PPT (2)用户要求做英语课PPT、英语公开课、初中英语教学PPT (3)用户提到"班会"或"英语课"与PPT结合
- References files that are not bundled: references/class班会-topics.md
- 1 referenced file(s) missing: references/class班会-topics.md
- Does not say what the result is
clawhub · ↓ 802
Integrationwill not run
本地 OCR / 图表识别技能。对本机图片做离线文字识别,支持截图、文档图片、扫描件、表格/图表图片中的中英文文本与图表数据提取;默认不上传文件、不依赖外部 API。
- References files that are not bundled: scripts/local_ocr.py, scripts/chart_ocr.py, scripts/regression.py
- 4 referenced file(s) missing: scripts/local_ocr.py, scripts/chart_ocr.py, scripts/regression.py
- Does not say what the result is
clawhub · ↓ 800
AnalyzerGitHubwill not run
LLM驱动的每日股票分析系统完整版 v2.1。支持A股/港股/美股智能分析、决策仪表盘、大盘复盘、板块分析、Agent问股、多渠道推送。提供技术面+基本面综合分析。触发词:股票分析、分析股票、每日分析、大盘复盘、板块分析、问股。
- References files that are not bundled: LICENSE
- 11 top-level sections: this looks like several domains in one skill
- 1 referenced file(s) missing: LICENSE
clawhub · ↓ 850
Integrationwill not run
DanceTempo / DanceTech Protocol — full repo context via public/llm-full.txt (or /llm-full.txt), tribal debugging via CLAWHUB.md, Tempo + MPP/x402 patterns, dance-extras live routes
- References files that are not bundled: scripts/build-llm-full.mjs
- 15 top-level sections: this looks like several domains in one skill
- 1 referenced file(s) missing: scripts/build-llm-full.mjs
clawhub · ↓ 823
Referencewill not run
系统思维技能,让 AI 具备分析复杂系统的能力
- References files that are not bundled: self.stocks
- 1 referenced file(s) missing: self.stocks
- Does not say what the result is
clawhub · ↓ 800
Integrationwill not run
Enterprise-grade agentic document processing API. Accurately extracts key fields and line items from invoices, receipts, orders and more across 10+ file formats, with confidence sc
- References files that are not bundled: ../../refers/backend/flows/ADP-Flow-Index.md, ../../refers/backend/flows/09-Doc-Extraction-Flow.md, ../../refers/backend/flows/08-Doc-Recognition-Flow.md
- Safety rules and hard prohibitions inside a skill: they belong in the system prompt, here they protect nothing
- 13 top-level sections: this looks like several domains in one skill
clawhub · ↓ 800
Referencewill not run
Ollama 模型迁移技能。用于将 Ollama 模型从 C 盘迁移到其他磁盘,释放 C 盘空间。使用场景:(1) C 盘空间不足需要迁移模型,(2) 更换模型存储位置,(3) 检查 Ollama 模型占用空间,(4) 验证迁移后模型可用性
- References files that are not bundled: scripts/verify_ollama.py
- 1 referenced file(s) missing: scripts/verify_ollama.py
- Does not say what the result is
clawhub · ↓ 778
IntegrationNotionwill not run
Notion via notion-cli — a Rust CLI + MCP server for Notion API 2025-09-03+. Three-tier agent integration (read-only default, opt-in runtime writes, opt-in admin lifecycle) with rat
- References files that are not bundled: ../docs/runtime-samples/, ../docs/cookbook/, ../docs/runtime-samples/hermes-profile.sample.yaml
- 14 top-level sections: this looks like several domains in one skill
- 5 referenced file(s) missing: ../docs/runtime-samples/, ../docs/cookbook/, ../docs/runtime-samples/hermes-profile.sample.yaml
clawhub · ↓ 822
Generatorwill not run
Expert in designing and building autonomous AI agents. Helps with agent architecture, tool integration, memory systems, planning strategies, and multi-agent orchestration. Use when
- References files that are not bundled: **kwargs
- 10 top-level sections: this looks like several domains in one skill
- 1 referenced file(s) missing: **kwargs
clawhub · ↓ 848
Procedurewill not run
Rank upcoming trade shows by how many of your competitors are exhibiting there. "Which shows are my competitors at?" / "竞争对手去哪些展会" / "Auf welchen Messen sind meine Wettbewerber?" /
- References files that are not bundled: url
- Safety rules and hard prohibitions inside a skill: they belong in the system prompt, here they protect nothing
- 1 referenced file(s) missing: url
clawhub · ↓ 906
Analyzerwill not run
Review, audit, coach, and extract PRISMA 2020 reporting compliance for systematic reviews, meta-analyses, protocols, reviewer comments, and draft manuscript sections. Use when chec
- References files that are not bundled: assets/source-docs/PRISMA_2020_checklist.docx, assets/source-docs/PRISMA_2020_expanded_checklist.pdf
- 11 top-level sections: this looks like several domains in one skill
- 2 referenced file(s) missing: assets/source-docs/PRISMA_2020_checklist.docx, assets/source-docs/PRISMA_2020_expanded_checklist.pdf
clawhub · ↓ 876
Procedurewill not run
Wires a real microphone through an AI brain (STT → LLM → TTS) and routes the output to a virtual audio cable so apps like Google Meet hear the processed voice. Use when building a
- References files that are not bundled: scripts/01_list_devices.js, references/architecture.md, references/step_by_step.md
- 4 referenced file(s) missing: scripts/01_list_devices.js, references/architecture.md, references/step_by_step.md
- Does not say what the result is
clawhub · ↓ 875
GeneratorWordPDFwill not run
政府公文格式文档生成器。支持所有常见文档格式输入(.md .txt .docx .pdf .html .rtf),输出符合国家标准的.docx公文格式。使用场景:产业调研报告、政策分析报告、工作报告等需要严格按照政府公文格式排版的文档。
- References files that are not bundled: url
- 1 referenced file(s) missing: url
- Does not say what the result is
clawhub · ↓ 822
Procedurewill not run
AgentMM memory & log management skill — gives AI agents persistent memory storage and structured logging. Use when the user asks to remember information, recall memories, or record
- References files that are not bundled: scripts/agentmm
- Safety rules and hard prohibitions inside a skill: they belong in the system prompt, here they protect nothing
- 1 referenced file(s) missing: scripts/agentmm
clawhub · ↓ 822
Procedurewill not run
Универсальный AI-скилл для путешествий — визы, новости, RSS, праздники, валюты с letundra.com. Используй when: пользователь спрашивает о визах, новостях, праздниках, курсах валют.
- References files that are not bundled: link
- 1 referenced file(s) missing: link
- Does not say what the process needs to start
clawhub · ↓ 798
Procedurewill not run
甲方首席防坑官。识别乙方套路、规避合作风险,输出可落地的全链路执行方案。触发:'帮我理理需求''怎么防坑''怎么落地''报价合理吗''验收标准''供应商选型'。不适用于纯技术实现或乙方内部管理。
- References files that are not bundled: references/industries/, references/industry-adaptation.md, references/participant-risk-map.md
- The response is described with custom markup (33 tags): a typed call is more reliable
- 9 referenced file(s) missing: references/industries/, references/industry-adaptation.md, references/participant-risk-map.md
clawhub · ↓ 1 174
Procedurewill not run
每日从多个权威新闻来源(BBC RSS、南华早报、36氪、TechCrunch、The Verge、Microsoft/NVIDIA官方博客等)抓取最新资讯,自动生成四板块新闻日报:国际重要新闻、中国重要新闻、AI重要新闻、AI科技巨头动态。输出带原文链接的Markdown文件。当用户需要生成每日新闻摘要、科技资讯日报、AI行业动态汇总、国际时事简报,或设置
- References files that are not bundled: URL
- 1 referenced file(s) missing: URL
- Does not say what the result is
clawhub · ↓ 797
Generatorwill not run
功能测试用例智能生成工具。用于根据需求文档、接口定义、数据库表结构、现有用例参考等输入,
自动生成标准格式的测试用例文档(Excel格式)。适用于:(1) 解读需求文档生成测试点 (2)
基于接口定义生成API测试用例 (3) 根据数据库表设计验证数据测试 (4) 参考现有用例生成
相似模块用例 (5) 批量生成测试用例并导出Excel
- References files that are not bundled: references/testcase_template.xlsx
- 1 referenced file(s) missing: references/testcase_template.xlsx
- Does not say what the result is
clawhub · ↓ 796