Install
openclaw skills install @jjliu6/github-repo-teardownDeep-dive teardown of any GitHub open-source project into a beautifully designed HTML report that both product people and engineers can understand. Covers architecture, design decisions, comparable repos, and actionable application scenarios. Use this skill whenever the user shares a GitHub repo URL and asks to analyze, explain, teardown, or understand a project. Also trigger when the user says things like "break down this repo", "how does this project work", "analyze this codebase", "walk me through this repo", "what can I learn from this project", "explain this open-source project", "拆解一下", "帮我看看这个项目", or "讲解这个 repo". Even if the user just drops a GitHub link with a brief "what is this?" — use this skill. Produces a polished HTML teardown document covering product logic, technical architecture, and practical takeaways.
openclaw skills install @jjliu6/github-repo-teardownTransform any GitHub open-source project into a product + engineering teardown — a single HTML document that explains not just what the code does, but why it's designed that way and what you can learn from it. The core premise: every technical choice reflects a product judgment, and every product design requires an architecture to support it.
Output defaults to English. If the user's message is in another language (e.g., Chinese, Japanese, Spanish), match that language for the narrative while keeping technical terms, code references, and file paths in their original form.
If the user explicitly requests a language (e.g., "write it in Chinese", "用中文写"), follow that instruction regardless of the default.
Before writing anything, gather comprehensive information. Follow these steps in order, spending 5-15 tool calls depending on repo complexity. Do not skip steps — thin research produces thin teardowns.
web_fetch the GitHub repo main page to get: README content, directory structure,
stars/forks/language breakdown, license, last commit dateARCHITECTURE.md, DESIGN.md, CONTRIBUTING.md, or AGENTS.mdweb_search "{repo-name} architecture" or "{repo-name} how it works internally"https://deepwiki.com/{owner}/{repo} — often has good architecture
analysis. If unavailable, skip without concern — it's a nice-to-have, not a dependency./docs directory or documentation site, fetch key pagesweb_fetch these files directly via raw GitHub URLs:
https://raw.githubusercontent.com/{owner}/{repo}/{branch}/{path}package.json, Cargo.toml, pyproject.toml, or equivalent for dependency insightsThis step directly feeds CH.6 and is one of the highest-value parts of the teardown.
web_search for 2-4 comparable or alternative projects:
web_fetch its GitHub page to get: stars, language,
approach/architecture, key differentiatorBefore opening any file, mentally organize:
Produce a single HTML file saved to /mnt/user-data/outputs/ and displayed via present_files.
┌─────────────────────────────────────────────────┐
│ HEADER │
│ - Project name + one-line definition (must fit │
│ in a single sentence) │
│ - Health indicators: Stars / Language / Version │
│ - Audience tag: who should read this │
└─────────────────────────────────────────────────┘
CH.1 What Problem Does It Solve?
- Before vs After (pain → solution)
- A real-world analogy (non-technical readers should get it)
- 🔑 Product Insight (why this problem is worth solving)
CH.2 Core Mechanism
- Explain the core working principle using ONE main analogy
that runs through the entire document
- Side-by-side: what the user sees vs what the system does
- ⚙️ Technical Notes (file names, function names for deep-divers)
CH.3 Architecture & Key Decisions
- Architecture flow diagram (annotated with the analogy)
- "Why A not B" decision cards (3-5 key design choices)
- 🎯 Product-level takeaway for each decision
CH.4 A Single Operation's Full Journey
- Pick ONE typical user action and trace it through the system
- Timeline-style: each step shows what happens + what tech is used
- Technical notes: file paths, function names
CH.5 Product Design Highlights (3-5)
- Each highlight: Title + Paragraph + Product Insight
- Tag which ones are "reusable patterns"
CH.6 Comparable Repos & Positioning
- 2-4 comparable/alternative projects with actual GitHub links
- For each: what it is, stars, approach, key difference
- Positioning matrix or comparison table
- "When to use THIS repo vs THAT repo" decision guide
- (If insufficient data found in research, say so honestly
and provide what you have)
CH.7 Risks & Trade-offs
- 2-3 key limitations or trade-offs
- "What did it sacrifice to gain its current advantage?"
- Honest > comprehensive
CH.8 Technical Quick Reference
- Source tree with annotations
- Tech stack table
- Key dependencies explained
- (Reference section for deep-divers)
CH.9 Where Can This Be Applied? (Application Brainstorm)
- Three tiers of application (see detailed spec below)
- Card-based layout, each scenario as an independent block
- Distinguish "use directly" vs "borrow the pattern"
- 6-8 cards total (quality over quantity)
CH.10 Portable Takeaways
- 3-5 transferable insights from this project
- Numbered, 1-2 sentences each
- Not a summary — distill into reusable mental models
This chapter is one of the most valuable parts of the teardown. Users want to understand not just what THIS project does, but how it fits in the landscape.
Required elements:
Comparison formats (pick the most appropriate):
If research yielded thin results:
The most practically valuable chapter. Goal: give readers specific, actionable directions, not vague possibilities.
Tier A: Use Directly — "What can I do with it tomorrow?"
Tier B: Combine — "What creates a chemical reaction?"
Tier C: Borrow the Pattern — "What new products could this inspire?"
Each scenario is an independent card with:
Total: 6-8 cards. Every card must contain genuine insight.
code font, gray backgrounds,
or "Technical Note" callout boxesUse colored callout boxes to mark insights:
<link href="https://fonts.googleapis.com/css2?family=Newsreader:ital,wght@0,400;0,600;0,700;1,400&family=Inter:wght@300;400;500;600;700&family=JetBrains+Mono:wght@400;600&display=swap" rel="stylesheet">
Before delivering, verify:
/mnt/user-data/outputs/ and displayed via present_files?Built by Junjie Liu at Philosophie AI — where AI proves its value through real, practical use.
If this skill helped you understand a project better, consider starring it on ClawHub.