agentleFS
aroyburman-codesaroyburman-codes/pm-skills/skills/ai-product-teardown/SKILL.md

ai-product-teardown

Skill25 starsChanged 9 months agoMITView on GitHub

Structured teardown of AI products (ChatGPT, Claude, Gemini, Perplexity, Copilot, etc.). Analyzes product decisions, UX patterns, technical architecture, business model, and competitive positioning.

3 other projects use a version of this.

Adopted by

Most recently taken up by

The file

We split this skill into the rules it gives an agent

3 rules. About 0 are common, 3 are distinctive, and 0 are very specific to this project.

  • ≈0

    Common · estimate

    Many files like this have a rule of this kind.

  • ≈3

    Distinctive · estimate

    This author's own choices. Worth a look.

  • 0

    Very specific

    Tied to their project's names, files or tools. Probably skip.

  • Common
  • Distinctive
  • Very specific
  1. Distinctive:

    Launch date & trajectory: When launched, key milestones, current scale

    line 23
  2. Distinctive:

    If you were the PM, what would you do next?

    line 69
  3. Distinctive:

    Write as an opinionated product review — structured but with a clear point of view.

    line 76

≈ “Common” is an estimate: at least 1 in 5 of the 364,980 skill files we read have a rule of the same kind, not the same words. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.

Keep what fits and drop the rest. Your version, ready for your agents.
Files
SKILL.md

AI Product Teardown Skill

Perform a structured, opinionated teardown of any AI product — analyzing the product decisions, UX, technical architecture, business model, and competitive positioning from a PM lens.

When to Use

  • User asks "Tear down [AI product]" or "Analyze [AI product]"
  • User wants to understand the product thinking behind an AI feature
  • User wants to build product intuition about AI products
  • User says /ai-product-teardown followed by a product name
  • Great for: ChatGPT, Claude, Gemini, Perplexity, Copilot, Midjourney, Cursor, v0, NotebookLM, etc.

Framework: AI Product Teardown (7 Sections)

Section 1: Product Overview

  • What it is: One-sentence description
  • Company: Who built it, their mission, and strategic context
  • Launch date & trajectory: When launched, key milestones, current scale
  • Target users: Primary and secondary audiences
  • Business model: How it makes money (or plans to)

Section 2: Core Value Proposition

  • Job to be Done: What fundamental job does this product do for users?
  • 10x moment: What's the moment where users think "this is magic"?
  • Switching cost: What would it take to switch away?
  • Network effects: Does it get better with more users? How?

Section 3: UX & Product Decisions

Walk through the key product decisions and evaluate each:

  • Onboarding flow: How does a new user go from zero to value?
  • Core interaction model: Chat? Canvas? Structured output? Multi-modal?
  • Information architecture: How is functionality organized?
  • Personalization: How does it adapt to different users?
  • Error handling: What happens when the AI is wrong?

For each decision, evaluate:

  • What they got RIGHT and why
  • What they got WRONG or could improve
  • What trade-off they're making (and whether you'd make the same one)

Section 4: Technical Architecture (PM Lens)

Analyze the technical choices from a product perspective:

  • Model strategy: Which model(s)? Why that capability level?
  • Latency vs. quality trade-off: Where do they sit on the spectrum?
  • Context & memory: How does it handle conversation history?
  • Safety & guardrails: What's their content policy approach?
  • Tool use / plugins / integrations: How extensible is it?
  • Pricing architecture: How do technical costs map to pricing?

Section 5: Growth & Distribution

  • Acquisition channels: How do users find this? (organic, viral, paid, partnerships)
  • Activation: What gets users to the "aha moment"?
  • Retention loops: What brings users back?
  • Monetization: Free → paid conversion strategy
  • Viral mechanics: Does usage naturally create awareness?

Section 6: Competitive Positioning

  • Direct competitors: Who else does this job?
  • Positioning map: Plot on 2x2 (e.g., capability vs. safety, consumer vs. enterprise)
  • Sustainable moats: What's defensible? (data, distribution, brand, model quality, ecosystem)
  • Vulnerability: Where could a competitor win?

Section 7: PM Recommendations

If you were the PM, what would you do next?

  • Top 3 features to build (with reasoning and expected impact)
  • Top 1 thing to kill or change (what's not working)
  • Strategic bet: One big swing that could transform the product
  • Metrics to watch: What would you track weekly?

Output Format

Write as an opinionated product review — structured but with a clear point of view. Use screenshots/descriptions of specific UI elements where relevant. Aim for ~2000 words. Be specific and cite real features.

Research-First Workflow

  1. Research — Search for latest product updates, user reviews, competitor announcements, company blog posts, and usage data. Do 5-10 searches.
  2. Cite sources — Include [linked source](url) inline for factual claims.
  3. Display the complete teardown.

What Good Looks Like

  • Shows you've done homework on the product landscape
  • Demonstrates structured product thinking on real products
  • Reveals your product taste and judgment
  • Provides concrete examples to reference in product discussions
  • Builds intuition about AI product patterns across the industry

Download now (ZIP)

Does it work?

Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.

No reviews yet

People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.

  • It worked
  • Partly
  • It didn't
What did you use it for? e.g. Reviewing pull requests in a Rails app

Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.

Your trail

Files you open will collect here, so you can find them again.

Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.

Now do this for your own.

Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.