skillhub-mcp-server
Provides access to a Docker image optimization skill, enabling AI agents to discover and use best practices for minimizing Docker image size.
Click on "Install Server".
Wait a few minutes for the server to deploy. Once ready, it will show a "Started" state.
In the chat, type
@followed by the MCP server name and your instructions, e.g., "@skillhub-mcp-serversearch for a skill to optimize docker images"
That's it! The server will respond to your query, and you can continue using it as needed.
Here is a step-by-step guide with screenshots.
skillhub-mcp-server
把 SkillHub 上 80+ 个实战技能 通过 MCP(Model Context Protocol)协议,自动暴露为任意 AI Agent 可一键调用的工具。 一份代码,同时服务 X-Agent AI MCP Hackathon 2026(开放创新赛道) 与 世界 AI 技能锦标赛 WAISC(自成长主题,补官方明确要求的 MCP 接入文件缺口)。
一、定位(Positioning)
SkillHub 已沉淀 80+ 个覆盖编程 / 办公 / 内容创作 / 数据分析 / 安全运维的技能(每个都是一份结构化的 SKILL.md:适用场景 / 工作流 / 示例 / 边界 / 技术底座)。但这些技能原本"只能被人读、不能被 Agent 调"。
skillhub-mcp-server 是它们的标准化 MCP 接入层:用零依赖的纯 Python 实现 MCP over stdio(JSON-RPC 2.0),让 Claude Desktop / Cursor / 任意自研 Agent 能通过标准 MCP 协议,检索、读取、并把任意技能自动包装成可调用的 tool。
Related MCP server: aiskillstore
二、场景(When to use)
Agent 在动手前先"查能力":与其让 LLM 凭空写一段可能出错的正则/ Dockerfile / 文献综述,不如先
search_skills("docker 镜像瘦身"),发现已有成熟技能,直接采纳。多 Agent 协作:一个"调度 Agent"通过
list_categories了解能力版图,把子任务分派给对应技能 tool。技能即服务(Skill-as-a-Tool):
skill_as_tool(slug)把任一个技能自动生成 MCP tool schema,是 SkillHub 技能接入 MCP 生态的标准姿势。WAISC 自成长主题契合点:server 暴露的技能目录可被上层 Agent 持续检索与复用,越用越懂"有哪些现成能力可用",本身就是一种"越用越懂你"的能力沉淀机制。
三、能力(Capabilities)
MCP 工具清单(4 个,tool schema 由 server.py 的 TOOLS 定义,自动随技能目录更新):
工具 | 作用 |
| 按关键词 / 类目 / 标签检索技能,返回 slug、名称、类目、标签、一句话描述,并标注是否已发布 |
| 读取某技能完整信息:frontmatter 元数据 + |
| 返回技能能力版图(类目:数量分布) |
| 把任一技能自动包装成可调用的 MCP tool schema(name=slug, description=技能描述, inputSchema 基于语义生成) |
零依赖:仅用 Python 标准库,Python 3.8+ 可直接运行,无
pip install。数据驱动:技能目录来自
manifest.json(由candidates/下全部SKILL.mdfrontmatter 自动生成),新增技能无需改代码。协议合规:实现 MCP 2024-11-05 的
initialize/tools/list/tools/call,含notifications/initialized、ping处理。
四、技术底座(Tech Stack)
传输:stdio(标准输入/输出),每行一个 JSON-RPC 2.0 消息,符合 MCP 标准传输约定;日志走 stderr 不污染协议流。
协议:
initialize协商protocolVersion=2024-11-05,声明capabilities.tools;tools/list返回工具定义;tools/call执行并返回content[].text。检索:
search_skills采用"英文/数字词整体 + 中文字串 2-gram 展开 + 命中数排序"的轻量相关度算法,对中文混合查询鲁棒(如 "docker 镜像 瘦身" 能正确命中docker-image-optimizer)。接入层:
skill_as_tool体现"Skill → MCP tool"的标准化映射,是 SkillHub 技能接入 MCP 生态的关键一环。
五、运行说明(Run)
# 1. 准备:确保 manifest.json 存在(已随仓库提供,含 80+ 技能元数据)
cd compete/x-agent-mcp
# 2. 直接运行(stdio 模式,等待 MCP 客户端连接)
python server.py
# 3. 自检:用内置最小客户端走通 initialize → tools/list → tools/call
python tests/test_client.py接入 Claude Desktop(claude_desktop_config.json):
{
"mcpServers": {
"skillhub": {
"command": "python",
"args": ["/绝对路径/compete/x-agent-mcp/server.py"]
}
}
}六、测试样例(Examples)
示例 1 — 检索技能(输入 → 输出)
→ tools/call: search_skills({"query": "docker 镜像 瘦身"})
← 命中 2 个技能(目录共 88 个):
- [✅已发布] docker-image-optimizer 《Docker 镜像瘦身专家》 [dev-programming] 把动辄 1GB+ 的镜像压到最小…
- [✅已发布] cicd-pipeline-builder 《CI/CD 流水线搭建与红灯排错》 [dev-programming] …示例 2 — 把技能包装成 tool(输入 → 输出)
→ tools/call: skill_as_tool({"slug": "docker-image-optimizer"})
← {
"name": "docker-image-optimizer",
"description": "把动辄 1GB+ 的镜像压到最小…",
"inputSchema": {"type":"object","properties":{"task":{"type":"string"}},"required":["task"]},
"annotations": {"category":"dev-programming","tags":["Docker","镜像优化",…]}
}示例 3 — 读取完整技能(输入 → 输出)
→ tools/call: get_skill({"slug": "secret-leak-scanner"})
← {"meta":{…frontmatter…}, "skill_md":"---\nname: secret-leak-scanner\n…全文…"}仓库根
tests/test_client.py可一键复现上述三个示例,作为"可运行、可验证"的提交证据。
Tool Schema Changelog
Recent tool additions, removals, and schema changes observed during successful MCP inspections. Dates show when Glama detected each change.
No tool schema history has been recorded yet.
This server cannot be installed
Maintenance
Resources
Unclaimed servers have limited discoverability.
Looking for Admin?
If you are the server author, to access and configure the admin panel.
Related MCP Connectors
Search & install 6,500+ AI agent skills from skills-hub.ai inside any MCP tool.
Governed AI agent skills — one library, distributed to devs and exposed to remote agents over MCP.
Free public MCP for AI agents — 193 tools, 44 workflows. No API key.
Sharebench — search & pull AI skills, agents, prompts & playbooks (SKILL.md) into any MCP client
Related MCP Servers
- AlicenseNot gradedqualityCmaintenanceA management toolkit for AI agent skills that provides an MCP server for search-first skill discovery and on-demand loading. It enables users to validate, organize, and serve standardized skills to MCP-compatible clients like Cursor and GitHub Copilot.409MIT
- AlicenseAqualityFmaintenanceAgent-first skill marketplace MCP server. AI agents discover, install, and share skills across 7 platforms via MCP protocol. 15 tools including skill search, download, upload, and agent discovery.184MIT
- AlicenseAqualityDmaintenanceUnified MCP and skill management gateway for AI agents, enabling tool discovery, installation, and sharing with 99% context token savings.819299Apache 2.0
- AlicenseNot gradedqualityAmaintenanceAn MCP server that lets AI agents search, view, install, and validate skills from the skillhub registry through tool calls.MIT
Latest Blog Posts
- Who's Calling? MCP Hosts Are an Identity Blind Spot (And the Spec Knows It)By Om-Shree-0709 on .mcpAgent IdentityOAuth 2.1
- Your AI Chatbot Just Exposed Your CEO's Salary to an InternBy Om-Shree-0709 on .Agent IdentityMCP SecurityOAuth Delegation
- Why MCP Servers Need Execution Sandboxing (And Why Your Current Stack Isn't Enough)By Om-Shree-0709 on .Agentic AiPrompt InjectionWebAssembly
MCP directory API
We provide all the information about MCP servers via our MCP API.
curl -X GET 'https://glama.ai/api/mcp/v1/servers/q15004040209-creator/skillhub-mcp-server'
If you have feedback or need assistance with the MCP directory API, please join our Discord server