GitHub Prior Art Research
GitHub Prior Art Research
Purpose
This skill activates when you ask questions about implementation approaches or tool selection. It guides Claude to research GitHub for proven solutions, popular libraries, real-world examples, and community discussions before formulating an answer.
When This Skill Activates
This skill automatically engages when your questions include patterns like:
- "How do I [implement/build/create/add] X?"
- "What's the best way to [solve/approach/handle] X?"
- "How should I structure/organize/design X?"
- "Which library/tool/framework should I use for X?"
- "What are people using for X?"
- "How might we [implement/architect] X?"
Research Process
1. Identify the Core Problem
Extract the key concept or task from your question. What's the actual problem you're solving?
2. Search Multiple Sources on GitHub
Use the WebSearch tool to find relevant information on GitHub:
Code Examples: Search for implementation patterns
- Example:
site:github.com "how to implement [X]" language:[relevant] - Look for well-maintained repos with multiple stars
Popular Repos: Find established solutions
- Search for repos that solve your problem
- Review their approach, architecture, and design decisions
Issues & Discussions: Learn from community problem-solving
- Search GitHub issues for discussions about similar challenges
- See what problems others encountered and how they solved them
Documentation: Find best practices and patterns
- Check README files and docs in relevant repos
- Look for architectural decisions and trade-offs explained
3. Synthesize Findings
Analyze what you discovered:
- What approaches are most common?
- What patterns do successful projects use?
- What trade-offs exist between different approaches?
- What mistakes do people make (from issues/discussions)?
4. Present Evidence-Based Answer
Propose solutions grounded in your research:
- Cite specific repos or discussions
- Explain why certain approaches work
- Mention alternatives and their trade-offs
- Point to real examples the user can study
Key Principles
- Always search before proposing: GitHub research informs every recommendation
- Cite sources: Include repo links or discussion references
- Show alternatives: Discuss different approaches and their trade-offs
- Learn from mistakes: Include common pitfalls found in issues/discussions
- Respect complexity: Acknowledge when multiple valid approaches exist
Example Usage
User asks: "How do I implement real-time updates in a React app?"
Skill activates because: The question matches "How do I implement [X]"
Claude's process:
- Searches GitHub for popular React real-time solutions (Firebase, Socket.io, etc.)
- Examines top repos and their architectural approaches
- Reviews issues discussing real-time update challenges
- Reads documentation explaining different patterns
- Proposes solution citing specific repos: "Based on popular approaches like [RepoA] and [RepoB], here are two main patterns..."
More from trevors/dot-claude
book-reader
Read and search digital books (PDF, EPUB, MOBI, TXT). Use when answering questions about a book, finding quotes or passages, navigating to specific pages or chapters, or extracting information from documents.
44using-git
Advanced git history operations and recovery — reflog for lost commits, cherry-pick with conflicts, revert on shared history, non-interactive `reset --soft` squash, aborting/recovering from rebases, stash across branches, pre-commit hook cycles. Prefers safe, non-interactive approaches. Skip for basic commit/push/branch/gitignore.
3analyzing-wav
Analyze WAV audio files for debugging TTS and audio pipelines. Use when checking audio quality, validating WAV format, inspecting waveform patterns, or diagnosing generated speech issues.
2github-prior-art
Search GitHub for prior art, code examples, and proven approaches before proposing solutions. Use when asked "how do I implement", "what's the best way to", "which library", "what tool should I use", or any implementation/tool selection question.
2executing-test-plans
Execute a structured test plan and emit a concise, comprehensive results document (JSON + Markdown). Use when running test plans, executing QA workflows, verifying acceptance criteria systematically, or producing test results documentation.
1downloading-hf-models
Get the correct command syntax for downloading models from HuggingFace. Use when downloading models, running "hf download", or pulling models from HuggingFace repos.
1