AI & me

How to Contribute to AIME Directory

October 8, 2025

You’ve been using AIME Directory. You’ve built collections, exported configurations, and sped up your development workflow. Now you want to give back. Maybe you’ve discovered an MCP that’s not in the directory. Or you’ve written an instruction file that’s made your team’s Copilot experience noticeably better. Perhaps you’ve crafted a prompt that saves you hours every week.

Perfect. Let’s talk about how to contribute to AIME Directory.

Why Contribute?

Before we dive into the how, let’s talk about the why. AIME Directory is only as useful as its content. Every MCP, instruction file, prompt, and configuration you see there was either:

  1. Automatically indexed from public registries (for MCPs)
  2. Manually curated and added by contributors

The more high-quality content we have, the more useful the directory becomes for everyone. Your contribution could save hundreds of developers hours of searching and configuration time. That’s a pretty good return on investment.

Plus, there’s a selfish reason: the more you contribute, the more you solidify your own understanding. Writing an instruction file forces you to articulate best practices clearly. Creating a reusable prompt makes you think about edge cases. Documenting an MCP setup ensures you’ll never forget how to configure it again.

What Can You Contribute?

AIME Directory accepts four types of contributions:

1. MCP Servers

Model Context Protocol servers extend Claude’s capabilities. If you’ve built an MCP server or found one that’s not in the directory, we want it. The directory automatically pulls from official registries, but there are always gems that aren’t registered yet.

Good MCP contributions include:

  • Clear purpose and use case
  • Working installation commands
  • Configuration examples
  • Proper tags for discoverability

2. Instructions

Framework-specific guidance files that teach AI assistants how to work with your tech stack. These go in .github/instructions/ and GitHub Copilot loads them automatically.

Good instruction contributions include:

  • Specific to a framework, language, or domain
  • Actionable guidelines, not generic advice
  • Real patterns that you use in production
  • Best practices backed by experience

3. Prompts

Reusable prompts for common development tasks. Code reviews, debugging sessions, documentation generation, architecture discussions - if you use it repeatedly, it belongs in the directory.

Good prompt contributions include:

  • Clear purpose and context
  • Structured format that produces consistent results
  • Placeholder sections for customization
  • Examples of when to use it

4. VSCode Configurations

Curated settings presets for GitHub Copilot and other AI tools. These merge intelligently when exported from collections.

Good config contributions include:

  • Focused on AI/Copilot features
  • Well-documented settings with comments
  • Real-world tested configurations
  • Clear use case (beginner, advanced, accessibility, etc.)

The Contribution Process

Contributing to AIME Directory is incredibly simple. No need to fork repositories, clone code, or set up development environments. Just click a button and propose your content.

The Easy Way: Propose New Content

Look for the plus icon (+) in the navbar at the top of any AIME Directory page. Click it, and you’ll be redirected to GitHub where you can create a new content proposal.

Here’s what happens:

  1. Click the “Propose new content” button (+ icon in navbar)
  2. You’re redirected to GitHub to create a new issue
  3. Fill in the proposal template with your content details
  4. Submit the issue for review
  5. Once approved, your content is integrated into AIME Directory

That’s it. No code changes, no pull requests, no technical setup required.

What to Include in Your Proposal

When you create a proposal issue, provide the following information based on what you’re contributing:

For all content types:

  • Title: Clear, descriptive name
  • Description: What it does and why it’s useful
  • Tags: 3-6 relevant tags for discoverability
  • Category (if applicable): Where it fits in the directory

Type-specific information:

Type-specific information:

For MCPs:

  • Repository URL
  • Installation command (e.g., npx -y your-package)
  • Type (stdio, sse, or http)
  • Configuration example
  • Use cases

For Instructions:

  • Target language/framework
  • Full instruction content (or attach as markdown)
  • When to use these guidelines
  • Real-world examples

For Prompts:

  • Prompt text (or attach as markdown)
  • When to use this prompt
  • Expected output format
  • Example use case

For VSCode Configs:

  • Settings JSON
  • What features the config enables
  • Target audience (beginner, advanced, etc.)
  • Usage tips

Example Proposal

Here’s what a good MCP proposal looks like:

Title: PostgreSQL Query Assistant MCP

Description: An MCP server that allows Claude to query PostgreSQL databases 
directly. Useful for data analysis, schema exploration, and debugging.

Type: MCP Server

Installation: npx -y @username/postgres-mcp

Configuration:
{
  "type": "stdio",
  "command": "npx -y @username/postgres-mcp",
  "env": {
    "DATABASE_URL": "postgresql://..."
  }
}

Tags: database, postgresql, sql, data-analysis, query

Repository: https://github.com/username/postgres-mcp

Use Cases:
- Query production databases for debugging
- Generate analytics reports
- Explore database schemas
- Test SQL queries

What Happens Next?

After you submit your proposal:

  1. Review: I review the proposal, usually within a few days
  2. Feedback: If anything needs clarification, I’ll comment on the issue
  3. Approval: Once approved, I integrate the content into the directory
  4. Live: Your contribution appears on aime.directory

You’ll be notified at each step through GitHub notifications.

Contribution Guidelines

To make the review process smooth, follow these guidelines:

Content Quality

Be Specific: “TypeScript Best Practices” is better than “Coding Standards”
Be Accurate: Test your configurations before submitting
Be Clear: Write like you’re explaining to a colleague
Be Concise: Provide enough detail, but respect people’s time

Formatting

Use Markdown Properly: Headers, lists, code blocks - format for readability
Include Examples: Show, don’t just tell
Tag Appropriately: 3-6 tags that accurately describe the content
Check Spelling: Proofread your content before submitting

Technical

Test Your Code: If it’s a configuration, make sure it works
Provide Sources: Link to official documentation where relevant
Keep It Updated: Use current versions and patterns
Respect Licenses: Only submit content you have the right to share

What Happens After You Contribute?

Once your pull request is merged:

  1. Your content appears in the directory immediately
  2. It’s indexed for search
  3. It shows up in filtered views based on tags
  4. Users can add it to collections and export it
  5. It helps hundreds of developers work more efficiently

You’ll also be listed as a contributor to the project. Not a bad addition to your GitHub profile.

Common Questions

Q: Do I need to know Git or coding to contribute?
A: No! Just click the “Propose new content” button and fill out a form. No technical skills required beyond using GitHub to submit an issue.

Q: Can I submit multiple items at once?
A: Yes, but create separate proposals for each item. This makes review easier and faster.

Q: What if I’m not sure my contribution is good enough?
A: Submit it anyway! The worst case is we’ll suggest improvements. The best case is it saves someone hours of work.

Q: Can I update existing content?
A: Absolutely. Click the propose button and mention you’re suggesting an update to existing content. Include what should change and why.

Q: What if the MCP I want to add is already in the directory?
A: Check first by searching. If it’s there but outdated or incomplete, propose an update with the new information.

Q: How long does review take?
A: Usually within a few days. Urgent fixes or high-value contributions might be reviewed faster.

Q: Can I contribute content in languages other than English?
A: Currently, the directory is English-only. If there’s demand for multilingual content, we might expand.

Q: What if my proposal is rejected?
A: I’ll always explain why and suggest how to improve it. Most rejections are due to missing information or unclear descriptions, which are easy to fix.

Special Cases

Proprietary MCPs

If you’ve built a proprietary MCP for your company that you can’t open-source, you can still add it to the directory with:

  • A clear description of what it does
  • Public documentation links (if available)
  • A note that it’s proprietary/private
  • Configuration examples (without sensitive data)

This helps others discover that the capability exists, even if they need to build their own implementation.

Company-Specific Instructions

If your instruction files contain company-specific patterns, you can:

  • Generalize them before submitting
  • Remove proprietary business logic
  • Keep the technical patterns and structures
  • Anonymize examples

The goal is to share the approach, not your company’s secrets.

The Bigger Picture

Every contribution to AIME Directory makes the AI development ecosystem a little bit better. You’re not just adding a file to a repository - you’re helping developers around the world work more efficiently.

Maybe someone discovers an MCP you contributed and it solves a problem they’ve struggled with for weeks. Maybe an instruction file you wrote helps a junior developer understand best practices. Maybe a prompt you shared saves someone from writing the same thing for the tenth time.

These are small acts with outsized impact.

So if you’ve got something valuable - an MCP, an instruction file, a prompt, a configuration - share it. The directory is better with your contribution, and you’ll be helping more developers than you realize.

Let’s Build Together

AIME Directory started as my personal solution to a problem I was experiencing. Now it’s growing into a community resource. The next phase depends on contributions from developers like you who see the value in curated, accessible AI development tools.

Ready to contribute? Click the “Propose new content” button (+ icon in the navbar), describe what you want to add, and submit. I’m looking forward to seeing what you contribute.

And if you have questions or ideas that don’t fit into a content proposal, open an issue on GitHub or reach out directly. Let’s make AIME Directory the best resource for AI-assisted development.

Happy contributing! 🌟