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.
GeneratorObsidianwill not run
Create and edit Obsidian Flavored Markdown with wikilinks, embeds, callouts, properties, and other Obsidian-specific syntax. Use when working with .md files in Obsidian, or when th
- References files that are not bundled: url
- 13 top-level sections: this looks like several domains in one skill
- 1 referenced file(s) missing: url
clawhub · ↓ 3 103
Generatorwill not run
Interview preparation system with company research, story building, and mock interview practice. Use when user mentions job interviews, interview prep, behavioral questions, salary
- References files that are not bundled: references/research.md, references/stories.md, references/mock-interviews.md
- Safety rules and hard prohibitions inside a skill: they belong in the system prompt, here they protect nothing
- 7 referenced file(s) missing: references/research.md, references/stories.md, references/mock-interviews.md
clawhub · ↓ 1 432 ★ 1
Template
Scaffold MCP server projects and baseline tool contract checks. Use for defining tool schemas, generating starter server layouts, and validating MCP-ready structure.
- Safety rules and hard prohibitions inside a skill: they belong in the system prompt, here they protect nothing
- Does not say what the result is
- No condition that starts the skill
clawhub · ↓ 2 553
Reference
将本地 Markdown/HTML 一键发布为在线链接,手机直接访问
- Does not say what the result is
- No condition that starts the skill
clawhub · ↓ 1 365 ★ 1
Templatewill not run
AI-powered China insurance actuarial pricing skill — uses the 4th Life Table (2025, effective 2026-01-01) and C-ROSS Phase II (Rules II 2024) framework. Calculates pure premium, re
- References files that are not bundled: references/life_table_2025.md, references/pricing_models.md, references/reserve_ifrs17.md
- 3 referenced file(s) missing: references/life_table_2025.md, references/pricing_models.md, references/reserve_ifrs17.md
- Does not say what the result is
clawhub · ↓ 1 738 ★ 1
Reference
- Prose only: no discrete steps
- Does not say what the result is
clawhub · ↓ 2 524
Integrationwill not run
CLI for AI agents to lookup country info for their humans. Uses REST Countries API. No auth required.
- References files that are not bundled: scripts/countries
- 1 referenced file(s) missing: scripts/countries
- Does not say what the process needs to start
clawhub · ↓ 2 556
Procedure
- Does not say what the result is
- No condition that starts the skill
clawhub · ↓ 2 517
Analyzer
Mint your unique AI agent avatar — CryptoPunks-style pixel art. Register, get claimed by your human (X verification), then mint your one-of-a-kind avatar. Use when an agent needs a
- Does not say what the result is
- No condition that starts the skill
clawhub · ↓ 2 514
GeneratorWord
专业的测试用例生成技能。支持文本/文档输入,自动生成高质量测试用例。内置依赖检测、质量评分、多格式导出、测试数据生成、用例评审、增量更新等高级功能。支持自定义 prompt 和配置。
- 11 top-level sections: this looks like several domains in one skill
- Does not say what the result is
- No condition that starts the skill
clawhub · ↓ 981 ★ 2
Procedure
通过 tccli 命令行管理腾讯云资源。
Load when: 用户要查询云服务器、管理 Lighthouse、购买域名、配置 DNS、
检查安全组、做安全巡检、部署应用、挂载云硬盘、部署 OpenClaw、
监控资源状态、SSL 证书检查、CAM 权限管理。
Covers: CVM、Lighthouse、CBS、VPC、DNSPod、SSL、CAM、Mon
- 11 top-level sections: this looks like several domains in one skill
- Does not say what the result is
- No condition that starts the skill
clawhub · ↓ 982 ★ 2
Analyzer
CANN 代码审查技能。用于审查 GitCode 上的 CANN 项目 PR。
当用户提到"审查 PR"、"代码审查"、"cann review"或提供 GitCode PR 链接时触发。
自动分析代码变更,检查内存泄漏、安全漏洞和可读性,生成结构化报告并发布评论。
- 12 top-level sections: this looks like several domains in one skill
- Does not say what the result is
- No condition that starts the skill
clawhub · ↓ 1 413 ★ 1
AnalyzerTelegramwill not run
Multiplayer text RPG with 10 playable worlds — play together in Telegram groups! Worlds: 笑傲江湖, Harry Potter, Warrior Cats, F15 Down, 龙虾三国, 倚天屠龙记, Game of Thrones, Lord of the Rings
- References files that are not bundled: assets/yumfu-logo.png
- Safety rules and hard prohibitions inside a skill: they belong in the system prompt, here they protect nothing
- 17 top-level sections: this looks like several domains in one skill
clawhub · ↓ 1 481 ★ 1
ProcedureDockerwill not run
Comprehensive ROS 2 engineering guide covering workspace setup, node architecture, communication patterns (topics/services/actions with QoS), lifecycle and component nodes, launch
- References files that are not bundled: references/*.md
- Safety rules and hard prohibitions inside a skill: they belong in the system prompt, here they protect nothing
- 1 referenced file(s) missing: references/*.md
clawhub · ↓ 980 ★ 2
Procedurewill not run
- References files that are not bundled: ../sci-data-extractor/, ../pubmed-search-skill/, ../sci-hub-search-skill/
- 3 referenced file(s) missing: ../sci-data-extractor/, ../pubmed-search-skill/, ../sci-hub-search-skill/
- No condition that starts the skill
clawhub · ↓ 1 370 ★ 1
Reference
- Does not say what the result is
- No condition that starts the skill
clawhub · ↓ 1 332 ★ 1
Generatorwill not run
Create new agent skills with proper structure, progressive disclosure, and bundled resources. Use when user wants to create, write, or build a new skill.
- References files that are not bundled: REFERENCE.md
- 1 referenced file(s) missing: REFERENCE.md
- Does not say what the result is
clawhub · ↓ 1 794 ★ 1
Procedurewill not run
KinemaClaw cross-platform Skill development and publishing specification. Defines the standard process for Codex and Claude plugin development, version management, marketplace inde
- References files that are not bundled: scripts/setup.reference.sh
- Safety rules and hard prohibitions inside a skill: they belong in the system prompt, here they protect nothing
- The response is described with custom markup (3 tags): a typed call is more reliable
clawhub · ↓ 1 502 ★ 1
Analyzerwill not run
Skill 标准化规范引擎。支持 R-01~R-26 规范审查(audit / create / update / refactor / bump / readonly 六模式),含权限扫描、数据目录合规检查、渐进式加载、LLM 二次筛分类。
- References files that are not bundled: references/*.md
- The response is described with custom markup (5 tags): a typed call is more reliable
- 1 referenced file(s) missing: references/*.md
clawhub · ↓ 2 898
Integrationwill not run
Deploy and manage serverless applications on Volcengine veFaaS. Use when the user wants to deploy web apps, agents, skills as APIs, tool pages, webhook functions, manage existing f
- References files that are not bundled: cookbooks/deploy-agent.md, cookbooks/publish-skill-as-api.md, cookbooks/generate-tool-page.md
- Safety rules and hard prohibitions inside a skill: they belong in the system prompt, here they protect nothing
- 4 referenced file(s) missing: cookbooks/deploy-agent.md, cookbooks/publish-skill-as-api.md, cookbooks/generate-tool-page.md
clawhub · ↓ 998 ★ 2
Procedure
AI周报总结
- Does not say what the result is
- No condition that starts the skill
clawhub · ↓ 1 836 ★ 1
Analyzerwill not run
- References files that are not bundled: ../../AGENT.SPEC.md, ../../SKILLS_MANAGEMENT.md
- 2 referenced file(s) missing: ../../AGENT.SPEC.md, ../../SKILLS_MANAGEMENT.md
- Does not say what the result is
clawhub · ↓ 1 403 ★ 1
Integration
- Does not say what the result is
- No condition that starts the skill
clawhub · ↓ 2 565
ProcedureGitHub
健康管理助手 - 基于十本畅销健康书籍共识的数据驱动健康管理系统。**快速暗号**: 健康记录、饮食分析、健康评分、周报、月报。**自然触发**: 记录饮食、查询营养、分析健康状况、补剂管理。支持多语言、多用户、时区自动识别。
- 17 top-level sections: this looks like several domains in one skill
- Does not say what the result is
- No condition that starts the skill
clawhub · ↓ 1 401 ★ 1
ProcedureGitHubwill not run
LDM OS installer and updater. Use when asked to install, update, or check status of LDM OS. Use when user pastes an install prompt mentioning wip.computer/install or ldm. Proactive
- References files that are not bundled: references/PRODUCT.md, references/SKILLS-CATALOG.md, references/COMMANDS.md
- Safety rules and hard prohibitions inside a skill: they belong in the system prompt, here they protect nothing
- The response is described with custom markup (12 tags): a typed call is more reliable
clawhub · ↓ 2 971