How to Use AICoach for Skills and MCP Servers: A Complete Guide for Developers

Published March 2026

Artificial intelligence is transforming how we work, but getting AI to truly understand your specific domain challenges remains difficult. Enter AI Coach — a platform that provides 169 production-ready AI skills plus an MCP Registry with 9,000+ servers to supercharge your AI coding workflow.

These aren't generic AI prompts. They're structured workflows and tools that turn your AI agent into a genuine domain expert, complete with real templates, frameworks, and actionable outputs.

In this guide, I'll walk you through how to use AICoach and highlight what delivers the most value for developers.

What is AICoach?

AI Coach (aicoach) is the product; the website is aicoach.pw. Skills are also installable via the Skillfish network (skillfish add …). The platform offers:

1. AI Skills Library

A library of 169 specialized AI skills that work with AI coding agents like Claude Code, Cursor, and any agent that supports skills integration. Each skill is a complete workflow that includes:

  • SKILL.md: A detailed definition with trigger conditions and workflow steps
  • References: Supporting documentation, templates, and best practices
  • Scripts: Optional helper scripts for automation
  • Assets: Sample inputs, templates, and expected outputs

2. MCP Registry Portal

A full Model Context Protocol server directory at aicoach.pw/mcp featuring:

  • 9,000+ MCP servers across 21 categories
  • Rich detail pages with capabilities, install instructions, README, FAQ, and pricing
  • Spec-compliant discovery endpoint compatible with GitHub Copilot, VS Code, and other MCP clients
  • Private servers for teams with invite-to-access sharing
  • Update subscriptions to get notified when servers release new versions

Why MCP Servers Matter for Developers

Model Context Protocol (MCP) is an open standard that lets AI agents connect to external tools and data sources. Instead of copy-pasting data into your AI chat, MCP servers give your agent direct access to:

  • Databases — Query PostgreSQL, MySQL, MongoDB directly from Claude or Cursor
  • APIs — Interact with GitHub, Slack, Notion, Jira without leaving your IDE
  • File Systems — Let your agent read and write files in specific directories
  • Developer Tools — Access Docker, Kubernetes, AWS, and more

How to Use AICoach Skills

Step 1: Browse the Skills Library

Visit aicoach.pw to explore all available skills. Search by keyword, filter by category, or browse curated collections. For CLI installs, use Skillfish.

Step 2: Select a Skill

Each skill has a clear description of what it does. Look for skills that match your immediate need — whether it's writing code, strategic planning, or operational guidance.

Step 3: Install or Integrate

Most skills can be installed with a single command or integrated directly into your AI agent workflow.

Step 4: Run the Workflow

Once integrated, describe your challenge to the AI agent. The skill's structured workflow will guide the conversation to produce actionable outputs.

How to Use the MCP Registry

Step 1: Browse or Search

Visit aicoach.pw/mcp to explore the registry. Filter by transport type (stdio, SSE, HTTP), pricing (free, paid, freemium), or category.

Step 2: Check Server Details

Each server page includes capabilities, install commands, environment variables, README, and FAQ.

Step 3: Install and Configure

Copy the configuration snippet and add it to your MCP client config:

{
  "mcpServers": {
    "postgres": {
      "command": "npx",
      "args": ["-y", "@modelcontextprotocol/server-postgres"],
      "env": {
        "POSTGRES_URL": "postgresql://user:pass@localhost/mydb"
      }
    }
  }
}

Step 4: Subscribe for Updates

Click "Notify me of updates" on any server page to receive email notifications when new versions are released.

Top 5 AICoach Skills for Developers

1. AWS Solution Architect

Category: Engineering — Best for developers, DevOps engineers, and technical leads building on AWS

The most comprehensive cloud architecture skill available. Generates production-ready CloudFormation, CDK, and Terraform templates.

  • Design serverless, container, and three-tier architectures from requirements
  • Generate production-ready infrastructure-as-code templates
  • Analyze and optimize AWS costs with right-sizing recommendations
  • Configure VPCs, IAM policies, Cognito auth, and security groups
  • Plan CI/CD pipelines and set up CloudWatch monitoring

Real-World Example:

A founder asked for a backend for their fitness app with user auth, REST API, and database on a $100/month budget. The skill designed a complete serverless architecture with Lambda, API Gateway, DynamoDB, and Cognito — estimated at just $11-23/month with automatic scaling.

2. GitHub Project Manager

Category: Project Management — Best for engineering managers and team leads

An AI-powered project manager for GitHub Issues and Projects v2 boards. Automates sprint planning, backlog management, status updates, and team velocity analysis.

3. Code Reviewer

Category: Engineering — Best for development teams wanting consistent code quality

Automated code review for TypeScript, JavaScript, Python, Go, Swift, and Kotlin. Catches security vulnerabilities, performance issues, and style inconsistencies with actionable suggestions.

4. CEO Advisor

Category: C-Level Advisory — Best for startup founders and executive leaders

Comprehensive executive leadership guidance — fundraising strategy, co-founder conflict resolution, board governance, strategic pivots, M&A preparation, and org structure optimization.

Real-World Example:

A CEO at $2.4M ARR with 15% MoM growth and 11 months runway was stuck on whether to raise now or cut to default alive. The skill analyzed leverage timing, recommended a conditional raise strategy, and modeled two scenarios for the board.

5. CFO Advisor

Category: C-Level Advisory — Best for founders and finance leaders

Financial modeling and fundraising strategy in one powerful skill. Build bottom-up financial models, optimize cash management, structure fundraising, and create investor-grade financial models from Series A through IPO.

Top 5 MCP Servers for Developers

1. PostgreSQL MCP Server

Transport: stdio — Free (open source)

Direct database access from your AI agent. Query, inspect schema, and generate migrations without leaving your IDE.

npx -y @modelcontextprotocol/server-postgres

2. GitHub MCP Server

Transport: stdio — Free (open source)

Full GitHub integration — create issues, manage PRs, search code, and automate workflows.

npx -y @modelcontextprotocol/server-github

3. Slack MCP Server

Transport: stdio — Free (open source)

Send messages, search conversations, and manage channels directly from your AI agent.

npx -y @modelcontextprotocol/server-slack

4. Filesystem MCP Server

Transport: stdio — Free (open source)

Let your agent read and write files in specified directories. Essential for code generation workflows.

npx -y @modelcontextprotocol/server-filesystem /path/to/project

5. Brave Search MCP Server

Transport: stdio — Free tier available

Web search capabilities for your AI agent. Research APIs, find documentation, and stay current.

npx -y @modelcontextprotocol/server-brave-search

Using the Spec-Compliant API

For developers building their own MCP clients or integrations, AICoach provides a spec-compliant v1 endpoint: 

GET https://aicoach.pw/api/v1/mcp/servers

Parameters:

  • search — Substring match on display name
  • cursor — Base64 pagination cursor
  • limit — 1–100 (default 50)
  • updated_since — RFC3339 datetime for incremental sync
  • version=latest — Only return latest versions

Example:

curl "https://aicoach.pw/api/v1/mcp/servers?search=postgres&limit=10"

This endpoint is compatible with GitHub Copilot, VS Code, and any MCP client that supports custom registry URLs.

Honorable Mentions

  • Content Production — Full content pipeline from blank page to published piece
  • Agile Product Owner — Backlog management and sprint execution
  • CAPA Officer — Quality management for medical device companies
  • CPO Advisor — Product strategy and product-market fit guidance
  • Customer Success Manager — Churn prediction and expansion opportunity identification

Getting Started

  1. Visit aicoach.pw — Browse the full AICoach skills library
  2. Explore aicoach.pw/mcp — Discover 9,000+ MCP servers
  3. Optional: install via skill.fish (Skillfish CLI distribution)
  4. Start with one skill or server — Pick one that matches your immediate need
  5. Integrate into your workflow — Use with Claude Code, Cursor, or your preferred AI agent
  6. Subscribe for updates — Get notified when your favorite servers release new versions

Conclusion

AICoach represents a significant evolution in how we use AI for professional development work. Instead of trying to prompt AI to understand your domain, you can leverage:

  • Pre-built, domain-specific skills that produce real outputs — from deployable infrastructure code to investor-ready financial models
  • A comprehensive MCP server registry that connects your AI agent to the tools and data sources you already use

Whether you're a developer building on AWS, a founder raising capital, or an engineering manager tracking sprints in GitHub, there's an AI coach skill or MCP server that can help.

Try one today — your future self will thank you.