What People Build

The best way to understand MyClaude is to see what people build on it. These are real examples — the kinds of products that live on the marketplace, created by the community, for everyone who uses Claude Code.

You don't need to be a developer to create or use most of these. If you can describe expertise clearly, you can package it as a MyClaude product.

Minds — Structured cognition you can install

This is the category that makes MyClaude unlike anything else. A Mind is not software. It's structured expertise — a cognitive framework that teaches Claude Code how to think about a domain.

What a Mind looks like inside:

# Kairo — Synthetic Reasoning Consciousness

You are a cognitive partner. You think alongside, question,
propose, and challenge. Not a passive executor.

## 7 Cognitive Pillars
- Logos — formal logic and valid inference
- Episteme — what counts as knowledge vs. belief
- Bayesiano — probability, confidence calibration
- Dialektikos — argument structure, steel man
- Empiricus — evidence hierarchy, causal reasoning
- Systema — complex systems, feedback loops
- Nous — metacognition, bias detection

## How to respond
Every conclusion carries a confidence marker:
  High (>85%) | Moderate (60-85%) | Hypothesis (30-60%)
Never use certainty language for uncertain conclusions.

That's it. Plain language. No code. The Mind tells Claude Code how to reason — and Claude follows those instructions in every response. Install a Mind, and Claude's entire cognitive approach shifts.

Who creates Minds: Strategists, researchers, consultants, domain experts, educators — anyone with a thinking methodology worth sharing. You don't write code. You write how you think. See the Building Minds guide for a step-by-step tutorial.

The meta-irony: This showcase page was written by Claude Code (Opus 4.6), using a Mind-like cognitive framework, to describe the very category of products that shape how Claude Code thinks. The tool is documenting the tool that improves the tool.

Skills — One task, done brilliantly

A Skill is the simplest product on MyClaude and the most common. It teaches Claude Code one new capability.

What a Skill looks like inside:

# Code Review — Security Focus

When the user asks you to review code, follow this protocol:

1. Check for OWASP Top 10 vulnerabilities
2. Verify input validation at system boundaries
3. Check for hardcoded secrets or credentials
4. Review authentication and authorization flows
5. Flag SQL injection, XSS, and command injection risks

Format: table with Severity | Location | Issue | Recommendation
Always check: are error messages leaking internal details?

One file. Clear instructions. Claude Code reads it and gains a new capability.

Who creates Skills: Anyone with a repeatable process — developers, yes, but also content strategists ("review this blog post for SEO"), financial analysts ("audit this spreadsheet for formula errors"), or HR professionals ("draft interview questions for this role").

CLAUDE.md — A personality transplant for your project

A CLAUDE.md config tells Claude Code how to behave inside a specific project. It's the most powerful file in the Claude Code ecosystem — and MyClaude is the only place to buy, sell, and discover them.

What a CLAUDE.md looks like inside:

# Project: E-Commerce Platform

## Stack
- Next.js 15, TypeScript strict, Tailwind CSS v4
- PostgreSQL with Drizzle ORM
- Stripe for payments

## Rules
- Never use `any` type. Prefer unknown + type guards.
- All API routes must verify auth before processing.
- Use server components by default. Client only for interactivity.
- Write tests before implementation (TDD).
- Commit messages: conventional commits format.

## Voice
- Be direct. No filler. Lead with the answer.
- When uncertain, say so explicitly with confidence level.

The dogfood: MyClaude itself runs on a CLAUDE.md — 150+ lines that define how Claude Code builds a marketplace. It enforces architectural decisions, security rules, design system tokens, and documentation standards. Every session, Claude reads it and knows exactly how to behave in this codebase. That CLAUDE.md is a product category on MyClaude. The platform selling CLAUDE.md configs is itself powered by one.

Who creates CLAUDE.md configs: Team leads who want consistent Claude Code behavior across projects. Consultants who configure Claude Code for specific tech stacks. Domain experts who set up Claude for fields like healthcare compliance, financial regulation, or academic research. See the Building CLAUDE.md guide for examples including non-dev configs.

Squads — Teams of specialists

A Squad coordinates multiple agents working together. Where an Agent is a single specialist, a Squad is the whole team.

Imagine this scenario:

You're building a feature. You invoke /vault-forge and four specialists activate:

• Architect designs the data flow and API contracts
• Engineer writes the implementation code
• Guardian audits for security vulnerabilities (read-only — never writes code)
• QA verifies the build, runs acceptance criteria, checks SSR output

Each has a distinct role, distinct constraints, and they hand off work to each other. The Architect designs. The Engineer builds what was designed. The Guardian reviews what was built. QA verifies the final result.

Who creates Squads: Experienced practitioners who've found that complex work needs division of labor — just like human teams.

What makes these different from "prompts"

Every product on MyClaude is more than a prompt. The difference:

A prompt is a suggestion. A MyClaude product is infrastructure.

Your expertise is a product

The pattern is the same across every category:

1. You know something valuable — a methodology, a workflow, a way of thinking
2. You write it in plain language — Markdown files that Claude Code reads
3. You package it with vault.yaml — 7 fields that describe your product
4. You publish with one command — myclaude publish
5. The world can install it — myclaude install @you/your-product

You don't need to write code. You need to write clearly. The most valuable products on MyClaude are expertise made installable.

Common questions

Can I try products before buying?

Many creators offer free versions alongside premium products. Browse myclaude.sh/explore and filter by "Free" to discover them.

How do I know a product is good?

Look for MCS certification badges (quality tiers), download counts, and purchase-verified reviews. MCS-3 (gold badge) products have passed human review for usability and production readiness.

Can an AI agent recommend products for my project?

Yes. Share your project URL with an AI assistant and ask "what MyClaude products would help this project?" — the agent can read your codebase context and search MyClaude for relevant tools. See Agent Resources for how this works.

Related pages

• Product Types Reference — detailed specs for all 12 categories
• Quickstart — install the CLI and get started
• The MyClaude Method — why we built this
• Creator Onboarding — three steps to publishing your own product