Which integrations are available for this server?

Allows searching for Docker-related agent skills and best practices, such as 'Docker security', and fetching their detailed instructions. Provides tools to search the SkillsMP marketplace for agent skills and fetch the raw SKILL.md instructions from GitHub repositories. Allows searching for Next.js-specific agent skills and best practices, such as 'Next.js caching', and fetching their detailed instructions.

How do I use skillsmp-mcp?

1. Click on "Deploy Server". 2. Wait a few minutes for the server to deploy. Once ready, it will show a "Started" state. 3. In the chat, type @ followed by the MCP server name and your instructions, e.g., "@skillsmp-mcp search the marketplace for a code review skill and fetch its SKILL.md" 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.

skillsmp-mcp

by Mekanikalistik

Overview Schema Source Related Servers Discussions Admin

Python

Local

SkillsMP MCP Server 🌐⚡

MCP Compatible Python Version Zero-Setup uvx License: MIT

A Model Context Protocol (MCP) server that connects your favorite AI agents and IDEs directly to the SkillsMP marketplace.

It enables LLMs (Claude, Gemini, GPT-4, Cursor, Antigravity, Windsurf) to autonomously search for curated agent skills, best practices, and workflows, then fetch the exact SKILL.md instructions on demand.

🛡️ Critical Directive: Read & Audit Before You Trigger

⚠️ "Knowing nothing about the skill you trigger is worse than blind vibe coding."

In the era of autonomous AI agents, Agent Skills are executable knowledge. They don't merely provide reference documentation—they define system heuristics, tool execution rules, architectural constraints, and operational boundaries for your AI agent.

Why Blind Skill Execution is Dangerous:

Vibe Coding vs. Skill Poisoning: If you blindly "vibe code", an LLM might generate a buggy snippet that fails a compiler or test run. But if you blindly inject an unvetted SKILL.md into your agent's runtime, you corrupt the agent's reasoning engine with outdated APIs, antipatterns, or conflicting directives that taint every subsequent task in your workspace.
The "Word Salad" Hazard: Many skills published on public marketplaces are generic boilerplate or copied API signatures. Injecting these wastes valuable context window tokens and degrades reasoning performance without offering actionable runbooks.
Permission & Action Exposure: Real agent skills often define terminal execution patterns, file modifications, or infrastructure interactions. Triggering a skill without knowing what it does compromises reproducibility and project safety.

📋 The 3-Point Audit Checklist Before Adopting Any Skill:

Inspect the SKILL.md: Always review the prompt triggers, step-by-step instructions, and expected tool calls before running it.
Filter Out Generic Boilerplate: If a skill just tells the model things standard foundation models already know (like basic syntax), discard it. Prioritize skills with novel runbooks, battle-tested workflows, or strict compliance steps.
Scope Responsibly: Prefer isolated, project-level scoping (.agents/skills/<name>/) instead of universally polluting your global agent configuration.

Related MCP server: SkillFlow MCP Server

🤖 1-Prompt Auto-Install for Your AI Agent

Don't want to configure JSON files manually? Copy and paste this prompt directly into your AI assistant (Google Antigravity, Cursor, Claude, Windsurf, or Roo Code):

Please configure the SkillsMP MCP server and its Skill Scout rule from https://github.com/Mekanikalistik/skillsmp-mcp for me:
1. Add "skillsmp" to my MCP server configuration using command "uvx" with args ["--from", "git+https://github.com/Mekanikalistik/skillsmp-mcp.git", "skillsmp-mcp"] and env {"SKILLSMP_API_KEY": "PASTE_KEY_HERE"}.
2. Download SKILL_SCOUT_PROTOCOL.md from the repo and save it to my agent rules/skills directory so you know how to scout, filter out "word salad", and ask for confirmation before installing skills.
3. Prompt me to paste my API key from https://skillsmp.com if I haven't already.

⚡ Zero-Setup Execution via `uvx` (Manual Setup)

No manual git clone, virtual environments, or dependency management required! If you have uv installed, you can run the server directly:

uvx --from git+https://github.com/Mekanikalistik/skillsmp-mcp.git skillsmp-mcp

🔌 One-Click Client Configuration

Add this server to your AI tool of choice. Just make sure to provide your SKILLSMP_API_KEY.

1. Claude Desktop

Add to claude_desktop_config.json:

macOS: ~/Library/Application Support/Claude/claude_desktop_config.json
Windows: %APPDATA%\Claude\claude_desktop_config.json

{
  "mcpServers": {
    "skillsmp": {
      "command": "uvx",
      "args": [
        "--from",
        "git+https://github.com/Mekanikalistik/skillsmp-mcp.git",
        "skillsmp-mcp"
      ],
      "env": {
        "SKILLSMP_API_KEY": "sk_live_skillsmp_your_api_key_here"
      }
    }
  }
}

2. Cursor

Add to your project's .cursor/mcp.json or Global Cursor Settings:

{
  "mcpServers": {
    "skillsmp": {
      "command": "uvx",
      "args": [
        "--from",
        "git+https://github.com/Mekanikalistik/skillsmp-mcp.git",
        "skillsmp-mcp"
      ],
      "env": {
        "SKILLSMP_API_KEY": "sk_live_skillsmp_your_api_key_here"
      }
    }
  }
}

3. Google Antigravity / Gemini IDE

Add to ~/.gemini/config/mcp_config.json:

{
  "mcpServers": {
    "skillsmp": {
      "command": "uvx",
      "args": [
        "--from",
        "git+https://github.com/Mekanikalistik/skillsmp-mcp.git",
        "skillsmp-mcp"
      ],
      "env": {
        "SKILLSMP_API_KEY": "sk_live_skillsmp_your_api_key_here"
      },
      "disabled": false
    }
  }
}

4. Windsurf / Roo Code / Cline

Configure as a Stdio MCP server:

Command: uvx
Arguments: ["--from", "git+https://github.com/Mekanikalistik/skillsmp-mcp.git", "skillsmp-mcp"]
Environment: {"SKILLSMP_API_KEY": "your_api_key_here"}

🔑 API Key Configuration

You can provide your SkillsMP API key in any of these ways:

Environment Variable (Best Practice): Pass SKILLSMP_API_KEY in your MCP client's env block.
Local Environment File: Copy .env.example or skillsmp.env.template to .env or skillsmp.env:
```
cp skillsmp.env.template skillsmp.env
```
Edit the file:
```
SKILLSMP_API_KEY=sk_live_skillsmp_your_actual_key
```
The server automatically detects .env, skillsmp.env, and ~/.gemini/config/skillsmp.env.

💡 Where to get a key? Register or sign in at skillsmp.com to generate your live API token.

🛠️ How It Works & Available Tools

               ┌───────────────────────┐
               │    LLM / AI Agent     │
               └───────────┬───────────┘
                           │ (stdio JSON-RPC)
               ┌───────────▼───────────┐
               │      skillsmp-mcp     │
               └───┬───────────────┬───┘
                   │               │
  search_skillsmp  │               │ fetch_skillsmp_skill
                   ▼               ▼
      ┌──────────────────────┐   ┌──────────────────────────┐
      │ skillsmp.com REST API │   │ raw.githubusercontent.com│
      │   (Search Marketplace)│   │   (Download SKILL.md)    │
      └──────────────────────┘   └──────────────────────────┘

The MCP server exposes two specialized tools:

1. `search_skillsmp`

Discovers skills in the SkillsMP marketplace using keywords.

API Endpoint: GET https://skillsmp.com/api/v1/skills/search
Authentication: Authorization: Bearer <SKILLSMP_API_KEY>

Parameters:

Parameter	Type	Required	Default	Description
`q`	`string`	Yes	—	Search query or keywords (e.g., `'Next.js caching'`, `'Docker security'`, `'Postgres RLS'`)
`limit`	`integer`	No	`20`	Maximum results to return (up to `100`)
`sortBy`	`string`	No	`"stars"`	Sort order: `"stars"` or `"recent"`
`category`	`string`	No	`None`	Filter by category slug (e.g., `'devops'`, `'data-ai'`, `'frontend'`)
`language`	`string`	No	`None`	ISO language code (e.g., `'en'`, `'zh'`)

Response Handling:
- 200 OK: Returns matching skills with titles, descriptions, stars, tags, and repository githubUrl.
- 401 Unauthorized: Returns error prompting the user to verify their API key.
- 429 Too Many Requests: Returns {"error": "DAILY_QUOTA_EXCEEDED"}.

2. `fetch_skillsmp_skill`

Retrieves the raw markdown instructions (SKILL.md) for any discovered skill.

Parameters:
Parameter
Type
Required
Description
githubUrl
string
Yes
The GitHub URL returned in search_skillsmp (e.g. https://github.com/user/repo/tree/main/skills/auth)
URL Transformation: The tool converts: https://github.com/{owner}/{repo}/tree/{branch}/{path} into: https://raw.githubusercontent.com/{owner}/{repo}/{branch}/{path}/SKILL.md and downloads the full markdown document.

🧠 Autonomous Agent Protocol (`SKILL_SCOUT_PROTOCOL`)

To help AI assistants use these tools effectively without bloating prompt context, we include a battle-tested rule: SKILL_SCOUT_PROTOCOL.md.

It instructs agents to:

Generate Specific Keywords: Avoid generic terms like "python"; use targeted long-tail phrases like "FastAPI JWT refresh rotation".
Apply the "Word Salad" Filter: Read the fetched SKILL.md and reject boilerplate, keeping only actionable, novel runbooks.
Consult & Scope: Ask the user before saving the skill to project-specific (.agents/skills/) or global directories.

💻 Manual / Local Development

If you wish to run or develop the server locally without uvx:

# Clone the repository
git clone https://github.com/Mekanikalistik/skillsmp-mcp.git
cd skillsmp-mcp

# Run with uv
uv run skillsmp-mcp

# Or standard pip
pip install -e .
python -m skillsmp_mcp.server

📄 License

Distributed under the MIT License.

Available Tools

2 tools

fetch_skillsmp_skillA

Fetches the full SKILL.md markdown content for a skill using its githubUrl returned from search.

Args: githubUrl: The githubUrl of the skill returned from search_skillsmp.

ParametersJSON Schema

Name	Required	Description	Default
`githubUrl`	Yes

Output Schema

ParametersJSON Schema

Name	Required	Description
`result`	Yes

TDQS

A4.2/5.0

Behavior3/5

Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?

Annotations are absent, so the description carries the full burden. It discloses the primary behavior (fetching full markdown content) but does not mention read-only nature, error handling, authentication, or rate limits. For a fetch operation, some additional context (e.g., no side effects, potential invalid URL behavior) would improve transparency, but the core behavior is adequately stated.

Agents need to know what a tool does to the world before calling it. Descriptions should go beyond structured annotations to explain consequences.

Conciseness5/5

Is the description appropriately sized, front-loaded, and free of redundancy?

The description is two concise sentences with a clear main purpose and an Args section that adds necessary context. No extraneous wording; it is front-loaded with the primary action and directly addresses the single parameter.

Shorter descriptions cost fewer tokens and are easier for agents to parse. Every sentence should earn its place.

Completeness4/5

Given the tool's complexity, does the description cover enough for an agent to succeed on first attempt?

With one parameter, no nested objects, and an output schema present, the description covers the essential usage: it explains what the tool does and what input it needs. It does not mention potential errors or prerequisites beyond the implicit need to have searched first, but given the simplicity, the description is nearly complete.

Complex tools with many parameters or behaviors need more documentation. Simple tools need less. This dimension scales expectations accordingly.

Parameters4/5

Does the description clarify parameter syntax, constraints, interactions, or defaults beyond what the schema provides?

Schema description coverage is 0%, but the description compensates by explaining the githubUrl parameter: 'The githubUrl of the skill returned from search_skillsmp.' This adds meaning beyond the bare schema (type string) and clarifies the parameter's origin and expected value, though it does not specify format or validation details.

Input schemas describe structure but not intent. Descriptions should explain non-obvious parameter relationships and valid value ranges.

Purpose5/5

Does the description clearly state what the tool does and how it differs from similar tools?

The description states a specific verb ('Fetches'), resource ('SKILL.md markdown content'), and input ('using its githubUrl returned from search'). It clearly distinguishes from the sibling search_skillsmp by indicating this is the follow-up fetch step, so an agent can readily identify its purpose.

Agents choose between tools based on descriptions. A clear purpose with a specific verb and resource helps agents select the right tool.

Usage Guidelines4/5

Does the description explain when to use this tool, when not to, or what alternatives exist?

The description explicitly ties usage to search results ('githubUrl returned from search_skillsmp'), implying it should be used after a search. It does not explicitly state when not to use it, but the complementary relationship with the sibling is clear. This is sufficient guidance for a simple two-tool workflow.

Agents often have multiple tools that could apply. Explicit usage guidance like "use X instead of Y when Z" prevents misuse.

search_skillsmpA

Searches the SkillsMP marketplace for agent skills, workflows, and prompts using keywords.

Args: q: The search query or keyword (e.g., 'automation', 'SEO', 'react'). limit: Number of results to return (default: 20, max: 100). sortBy: Sort order of the results ("stars" or "recent"). category: Optional slug to filter by category (e.g., 'data-ai', 'devops'). language: ISO code to filter by skill content language (e.g., 'en', 'zh').

ParametersJSON Schema

Name	Required	Default
`q`	Yes
`limit`	No
`sortBy`	No	stars
`category`	No
`language`	No

Output Schema

ParametersJSON Schema

Name	Required	Description
`result`	Yes

TDQS

A4.1/5.0

Behavior3/5

Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?

No annotations are provided, so the description carries the behavioral disclosure burden. It explains what is searched and lists parameters with defaults, but it does not mention authentication, rate limits, result behavior, or any side effects. For a read-only search tool, this is adequate but not deeply transparent.

Agents need to know what a tool does to the world before calling it. Descriptions should go beyond structured annotations to explain consequences.

Conciseness5/5

Is the description appropriately sized, front-loaded, and free of redundancy?

The description is concise and well-structured: a single opening sentence followed by a tight argument list. Every line adds value, and the key purpose is front-loaded.

Shorter descriptions cost fewer tokens and are easier for agents to parse. Every sentence should earn its place.

Completeness4/5

Given the tool's complexity, does the description cover enough for an agent to succeed on first attempt?

The description covers all parameters with useful examples and defaults, and an output schema exists so return-value details are not required. It is complete for a search tool, though it could briefly reference the sibling fetch tool for completeness.

Complex tools with many parameters or behaviors need more documentation. Simple tools need less. This dimension scales expectations accordingly.

Parameters5/5

Does the description clarify parameter syntax, constraints, interactions, or defaults beyond what the schema provides?

The schema has 0% description coverage, but the description fully compensates by documenting all five parameters: q with examples, limit with default and max, sortBy with valid values, category with slug examples, and language with ISO examples. This adds substantial meaning beyond the bare schema.

Input schemas describe structure but not intent. Descriptions should explain non-obvious parameter relationships and valid value ranges.

Purpose5/5

Does the description clearly state what the tool does and how it differs from similar tools?

The description clearly states the tool searches the SkillsMP marketplace for skills, workflows, and prompts using keywords. It names the specific resource and action, and the verb 'Searches' distinguishes it from the sibling fetch_skillsmp_skill.

Agents choose between tools based on descriptions. A clear purpose with a specific verb and resource helps agents select the right tool.

Usage Guidelines3/5

Does the description explain when to use this tool, when not to, or what alternatives exist?

The description implies the tool is for keyword-based discovery, but it does not explicitly mention when to use search versus fetch_skillsmp_skill, nor any exclusions or prerequisites. The usage context is clear but no direct guidance about alternatives is provided.

Agents often have multiple tools that could apply. Explicit usage guidance like "use X instead of Y when Z" prevents misuse.

Tool Schema Changelog

Recent tool additions, removals, and schema changes observed during successful MCP inspections.

2 tool updatesv0.1.0Sep 14, 2026
- First observedfetch_skillsmp_skill
- First observedsearch_skillsmp

Deploy Server

One endpoint for every MCP toolRayrun runs every MCP server behind one URL and holds the credentials. Each call is decided per client, tool, and argument, then recorded.ray.runAd

TDQS

A4/5.0

Scored 2026-09-14 17:35 across 2 tools

Disambiguation5/5

search_skillsmp is the discovery entry point, while fetch_skillsmp_skill retrieves a specific skill's markdown content using a githubUrl. Their inputs, outputs, and purposes are clearly distinct with no realistic overlap.

Naming Consistency4/5

Both tools use snake_case action-prefixed names, making the pattern mostly predictable. The minor inconsistency is that search_skillsmp names the marketplace while fetch_skillsmp_skill explicitly names the skill object.

Tool Count3/5

Two tools is borderline thin, but it is understandable for a narrow search-then-fetch marketplace lookup server. There are no separate management or browsing tools, so the count feels slightly under-scoped rather than bloated.

Completeness2/5

Search claims to cover skills, workflows, and prompts, but only skills have a fetch operation for their content. This leaves workflows and prompts as dead ends after search results are returned, which is a significant gap relative to the stated marketplace scope.

Maintenance

ActivityMaintained

ResponsivenessNo issues

Resources

GitHub Repository

Need Help?

Related Servers

Related MCP Connectors

skills-hub A
io.github.tinh2
Search & install 6,500+ AI agent skills from skills-hub.ai inside any MCP tool.
Autonomous Agents AI & Machine Learning Search
42
SuperSkill Registry A
com.onlyharness
Find, inspect and install reviewed reusable AI-agent capabilities from superskill.sh.
Agent Orchestration Developer Tools
122
emdly A
com.emdly
Search, install and submit Markdown skills for AI agents; private Brain files when signed in.
Knowledge & Memory Note Taking Documentation Access
91
Qumge Skills A
com.qumge
Search and install curated agent skills, plus bundles that get one job done in a single call.
Agent Orchestration Autonomous Agents Developer Tools
41

Related MCP Servers

skills-master-mcp
Developer Tools Coding Agents Autonomous Agents
adarc8
A
license
A
quality
D
maintenance
Connects AI coding agents to the SkillsMP marketplace, allowing users to search, read, and install over 8,000 community-made skills. It enables agents to gain new capabilities either through on-the-spot instruction or permanent installation without requiring an API key.
Updated 7 months ago (2026-02-16 23:50 UTC)
5
12 npm
10
MIT
SkillFlow MCP Server
AI & Machine Learning Developer Tools Search
rafsilva85
A
license
A
quality
D
maintenance
Connects AI coding agents to the SkillFlow marketplace to search, discover, and retrieve detailed information about agent skills. It enables users to browse trending skills, categories, and publisher data directly through MCP-compatible environments.
Updated 6 months ago (2026-03-26 18:33 UTC)
5
27 npm
1
MIT
skills.sh MCP Server
ANeuronI
A
license
A
quality
B
maintenance
Enables AI agents to autonomously search, evaluate, and install skills from the skills.sh catalog.
Updated 2 months ago (2026-07-19 16:04 UTC)
2
4
13 npm
ISC
skill-retrieval-mcp
Search Knowledge & Memory
JayCheng113
A
license
Not graded
quality
A
maintenance
Enables AI agents to search and retrieve over 89K skills on-demand at runtime, eliminating the need to manually install skills upfront.
Updated 19 days ago (2026-08-27 06:39 UTC)
125
MIT

SkillsMP MCP Server 🌐⚡

🛡️ Critical Directive: Read & Audit Before You Trigger

⚠️ "Knowing nothing about the skill you trigger is worse than blind vibe coding."

Why Blind Skill Execution is Dangerous:

📋 The 3-Point Audit Checklist Before Adopting Any Skill:

🤖 1-Prompt Auto-Install for Your AI Agent

⚡ Zero-Setup Execution via uvx (Manual Setup)

🔌 One-Click Client Configuration

1. Claude Desktop

2. Cursor

3. Google Antigravity / Gemini IDE

4. Windsurf / Roo Code / Cline

🔑 API Key Configuration

🛠️ How It Works & Available Tools

1. search_skillsmp

2. fetch_skillsmp_skill

🧠 Autonomous Agent Protocol (SKILL_SCOUT_PROTOCOL)

💻 Manual / Local Development

📄 License

Available Tools

Output Schema

TDQS

Output Schema

TDQS

Tool Schema Changelog

TDQS

Maintenance

Resources

Related MCP Connectors

Related MCP Servers

skills-master-mcp

SkillFlow MCP Server

skills.sh MCP Server

skill-retrieval-mcp

⚡ Zero-Setup Execution via `uvx` (Manual Setup)

1. `search_skillsmp`

2. `fetch_skillsmp_skill`

🧠 Autonomous Agent Protocol (`SKILL_SCOUT_PROTOCOL`)