link-project
Add project repository links to PROJECTS.md when user mentions completing, starting, or sharing a hands-on project. Use when user references external repos or implementation work.
When & Why to Use This Skill
This Claude skill automates the maintenance of your project portfolio by intelligently updating PROJECTS.md files. It monitors conversations for mentions of new or completed repositories, extracts relevant links, and organizes them into structured categories such as AI Engineering, Machine Learning, and Software Engineering, ensuring your documentation remains professional and current without manual editing.
Use Cases
- Scenario 1: Automatically adding a newly created GitHub repository link to your portfolio after announcing a project milestone in chat.
- Scenario 2: Categorizing various AI experiments (like RAG systems or LLM agents) into a structured markdown list to maintain an organized technical knowledge base.
- Scenario 3: Streamlining the documentation workflow by generating concise project descriptions and maintaining consistent formatting in project logs.
| name | link-project |
|---|---|
| description | Add project repository links to PROJECTS.md when user mentions completing, starting, or sharing a hands-on project. Use when user references external repos or implementation work. |
| allowed-tools | Read, Edit |
Link Project
When the user mentions a project repository they've built:
Instructions
Read the current PROJECTS.md file
Determine the appropriate category:
- Machine Learning - Classic ML algorithms, supervised/unsupervised learning
- Deep Learning - Neural networks, computer vision, NLP models
- AI Engineering - LLM applications, agents, RAG systems
- ML Systems & Production - MLOps, deployment, monitoring, data pipelines
- Software Engineering - General software projects, tools, libraries
- Other - Anything that doesn't fit above
Add the project link in format:
- [Project Name](repo-url) - Brief description of what it does/demonstratesMaintain organization:
- Group by category (create category header if it doesn't exist)
- Within each category, newest projects at the bottom
- Keep descriptions concise (one line)
Examples
## Deep Learning
- [CNN Image Classifier](https://github.com/user/cnn-classifier) - From-scratch CNN implementation for CIFAR-10
- [Transformer from Scratch](https://github.com/user/transformer) - Educational implementation of attention mechanism
## AI Engineering
- [RAG Document QA](https://github.com/user/rag-qa) - Retrieval-augmented generation system with vector database
Edge Cases
- If PROJECTS.md is empty: Create basic structure with category headers
- If project fits multiple categories: Use primary category or ask user
- If repo URL not provided: Ask user for the link
- If description not clear: Ask user what the project demonstrates