superpowers-mcp
An MCP server that makes superpowers skills available to any LLM that supports the Model Context Protocol.
README
superpowers-mcp
An MCP server that makes superpowers skills available to any LLM that supports the Model Context Protocol.
Superpowers is a skills library for Claude Code that enforces disciplined workflows -- TDD, systematic debugging, brainstorming, planning, and more. This server exposes those skills to any MCP-compatible client: Cursor, Windsurf, Gemini, Kilo Code, Claude Desktop, or your own app.
Quick Start
1. Clone and build
git clone https://github.com/erophames/superpowers-mcp.git
cd superpowers-mcp
npm install
npm run build
2. Run setup
node build/index.js
On first run in a terminal, the setup wizard clones the superpowers repository and saves the configuration.
3. Configure your MCP client
Add to your client's MCP server configuration:
{
"mcpServers": {
"superpowers": {
"command": "node",
"args": ["/absolute/path/to/superpowers-mcp/build/index.js"]
}
}
}
Replace /absolute/path/to/ with the actual path where you cloned the repository.
<details> <summary>Client-specific config file locations</summary>
| Client | Config Location |
|---|---|
| Claude Desktop | ~/Library/Application Support/Claude/claude_desktop_config.json |
| Claude Code | ~/.claude.json or project .mcp.json |
| Cursor | Cursor Settings > MCP |
| Windsurf | ~/.codeium/windsurf/mcp_config.json |
</details>
What It Exposes
The server registers skills through all three MCP primitives:
Tools
| Tool | Description |
|---|---|
list_skills |
List all available skills with descriptions and file lists |
use_skill |
Load a skill by name -- returns the full skill content as instructions to follow (optional guardrail enforcement) |
get_skill_file |
Load a supporting file from a skill (reference docs, prompt templates, scripts) |
recommend_skills |
Recommend top skills for a task using semantic ranking + workflow policy boosts |
compose_workflow |
Build an ordered multi-skill workflow for a goal |
validate_workflow |
Validate selected skills against required workflow guardrails |
semantic_search_skills |
Semantic search across SKILL.md and supporting files |
Prompts
Each skill is registered as an MCP prompt named superpowers:{skill-name}. Clients that support prompt selection will show these in their UI. Selecting a prompt injects the full skill content into the conversation.
Resources
Each skill's SKILL.md and supporting files are registered as resources:
superpowers://skills/brainstorming/SKILL.md
superpowers://skills/test-driven-development/testing-anti-patterns.md
superpowers://skills/systematic-debugging/root-cause-tracing.md
A resource template superpowers://skills/{skillName}/{fileName} enables dynamic access to any file.
Available Skills
| Skill | What it does |
|---|---|
brainstorming |
Collaborative design through questions, approaches, and incremental validation |
writing-plans |
Bite-sized implementation plans with exact file paths and TDD steps |
executing-plans |
Batch execution of plans with review checkpoints between tasks |
subagent-driven-development |
Fresh subagent per task with two-stage code review |
dispatching-parallel-agents |
Distribute independent tasks to concurrent agents |
test-driven-development |
RED-GREEN-REFACTOR cycle -- write failing test first, always |
systematic-debugging |
4-phase root cause analysis: investigate before you fix |
verification-before-completion |
Run the command, read the output, then claim the result |
requesting-code-review |
Dispatch code review and act on severity-categorized feedback |
receiving-code-review |
Technical rigor when receiving feedback -- verify, don't blindly agree |
using-git-worktrees |
Isolated git worktrees for parallel feature development |
finishing-a-development-branch |
Structured options for merge, PR, or cleanup when done |
writing-skills |
Framework for creating, testing, and deploying new skills |
using-superpowers |
How the skill system works and when to invoke skills |
Usage Examples
Once connected, ask your AI assistant:
- "What superpowers skills are available?" (calls
list_skills) - "Use the brainstorming skill to help me design a caching layer."
- "Load the TDD skill and follow it to implement this feature."
- "Read the anti-patterns file from the test-driven-development skill."
- "Recommend the best skills for implementing this MCP feature."
- "Compose a workflow for debugging flaky tests."
- "Validate this workflow: brainstorming, writing-plans, test-driven-development."
- "Search skills for root cause tracing techniques."
Configuration
Environment Variables
| Variable | Description |
|---|---|
SUPERPOWERS_SKILLS_DIR |
Override the skills directory path directly |
Skill Discovery Order
SUPERPOWERS_SKILLS_DIRenvironment variable- Saved directory from the setup wizard (persisted in
~/.config/superpowers-mcp) - Claude Code plugin cache (
~/.claude/plugins/cache/claude-plugins-official/superpowers/*/skills/)
Auto-Updates
When skills are sourced from a git repository, the server checks for updates once per day on startup and pulls new changes automatically.
Development
npm install
npm run build
npm test
| Script | Description |
|---|---|
npm run build |
Compile TypeScript |
npm run dev |
Watch mode compilation |
npm test |
Run all 58 tests |
npm run test:watch |
Watch mode tests |
npm start |
Run the server |
Architecture
src/
index.ts Entry point, stdio transport, setup and update orchestration
server.ts McpServer creation, composes all registrations
config.ts Persistent configuration
update.ts Daily auto-update check
git.ts Git operations (clone, pull, fetch) via execFile
cli/setup.ts Interactive setup wizard
skills/
types.ts Skill, SkillFile, SkillMetadata interfaces
discovery.ts Directory scanning, YAML frontmatter parsing, version resolution
tools/register.ts list_skills, use_skill, get_skill_file
prompts/register.ts One MCP prompt per skill
resources/register.ts Static resources per file + resource template for dynamic access
Tests
License
MIT
Recommended Servers
playwright-mcp
A Model Context Protocol server that enables LLMs to interact with web pages through structured accessibility snapshots without requiring vision models or screenshots.
Magic Component Platform (MCP)
An AI-powered tool that generates modern UI components from natural language descriptions, integrating with popular IDEs to streamline UI development workflow.
Audiense Insights MCP Server
Enables interaction with Audiense Insights accounts via the Model Context Protocol, facilitating the extraction and analysis of marketing insights and audience data including demographics, behavior, and influencer engagement.
VeyraX MCP
Single MCP tool to connect all your favorite tools: Gmail, Calendar and 40 more.
graphlit-mcp-server
The Model Context Protocol (MCP) Server enables integration between MCP clients and the Graphlit service. Ingest anything from Slack to Gmail to podcast feeds, in addition to web crawling, into a Graphlit project - and then retrieve relevant contents from the MCP client.
Kagi MCP Server
An MCP server that integrates Kagi search capabilities with Claude AI, enabling Claude to perform real-time web searches when answering questions that require up-to-date information.
E2B
Using MCP to run code via e2b.
Neon Database
MCP server for interacting with Neon Management API and databases
Qdrant Server
This repository is an example of how to create a MCP server for Qdrant, a vector search engine.
Exa Search
A Model Context Protocol (MCP) server lets AI assistants like Claude use the Exa AI Search API for web searches. This setup allows AI models to get real-time web information in a safe and controlled way.