SkillsPicker

npx add-mcp 'https://skillspicker.com/mcp'

MCP server

Let your coding agent find skills for the project it is working on. The agent derives tags from local files and sends those, so code stays on your machine.

Endpoint: https://skillspicker.com/mcp

Transport: Streamable HTTP. Auth: none. Read-only. Tools: recommend_skills, search_skills, list_skill_tags.

Add it

Claude Code

claude mcp add --transport http skillspicker https://skillspicker.com/mcp

Cursor (.cursor/mcp.json)

{ "mcpServers": { "skillspicker": { "url": "https://skillspicker.com/mcp" } } }

VS Code / GitHub Copilot (.vscode/mcp.json)

{ "servers": { "skillspicker": { "type": "http", "url": "https://skillspicker.com/mcp" } } }

Codex (~/.codex/config.toml)

[mcp_servers.skillspicker]
url = "https://skillspicker.com/mcp"

HTTP API

No MCP client? GET https://skillspicker.com/api/recommend?tags=nextjs,supabase,testing or ?q=your project in words returns the same free result as JSON. Spec: openapi.json.

The tools

recommend_skills takes tags (preferred), description or repo_url. With no input it returns a sample for a Next.js project. search_skills takes one query, for example "stripe webhooks". list_skill_tags takes no input and returns the tag names. Each search returns the match count, 2 free picks with install commands, and a link to the full curated list ($9 once).

Tag names: rules.json. Example prompt: "Use skillspicker to find skills for this repo. Derive tags from package.json first."