Back to search
competitive-teardown-marketplace
MarketplaceWorks with Claude CodeGitHub
Evidence-based competitive teardown skills and agents for product people.
Safety scan
A static text scan, not a guarantee. Select the badge to see what was scanned and every finding.
Badge for your README
Markdown
[](/listings/takethecan0lis--competitive-teardown)HTML
<a href="/listings/takethecan0lis--competitive-teardown"><img src="/api/listings/takethecan0lis--competitive-teardown/badge.svg" alt="findsafeskills safety rating" /></a>Install
Install instructions from the README
As written by the repo's author. findsafeskills has not verified these steps.
Needs a recent Claude Code. The repo has a marketplace file at its root.
/plugin marketplace add takethecan0lis/competitive-teardown /plugin install competitive-teardown@competitive-teardown-marketplace
What was scanned
The files the safety scan read, as written by the repository's author and shown as plain text. Lines with a finding are marked.
.claude-plugin/marketplace.json · Manifest · 14 lines
- {
- "name": "competitive-teardown-marketplace",
- "description": "Evidence-based competitive teardown skills and agents for product people.",
- "owner": {
- "name": "Drew Podwal"
- },
- "plugins": [
- {
- "name": "competitive-teardown",
- "source": "./",
- "description": "Evidence-based competitive teardowns of apps, websites, and desktop software, plus a market roll-up that finds your differentiator."
- }
- ]
- }
README.md · README · 75 lines
- # competitive-teardown
- This is a prompt. Pin a release tag and commit, read every file, scan it for injected text, and run it first where no credentials are present. Its agents fetch untrusted web pages and can run Bash. product-explorer drives your real phone: use a test account. Exports hold your personal data.
- Adopted, not measured: these routines are in use, but no result has been measured. Take the method, not a claim.
- ## What it is and who it is for
- A Claude Code plugin that runs evidence-based competitive teardowns of apps, websites and desktop software, then rolls several teardowns into a market view that points at a possible differentiator. It is for product managers, product owners, designers and researchers who want a competitor's product described neutrally, with every claim tagged by how it is known (observed, documented, reported or inferred).
- ## Install
- Needs a recent Claude Code. The repo has a marketplace file at its root.
- ```
- /plugin marketplace add takethecan0lis/competitive-teardown
- /plugin install competitive-teardown@competitive-teardown-marketplace
- ```
- ## First run
- Start with:
- ```
- /teardown <product>
- ```
- The skill asks a few charter questions (product, access, audience, focus, depth, deadlines), then works through desk research, exploration, analysis and packaging. Standard depth is the default.
- ## Skills and agents
- Skills:
- - teardown: orchestrates one teardown from charter to packaged knowledge base and summary.
- - product-explorer: maps every screen of a product with a strict create-and-cancel safety protocol.
- - help-center-extractor: turns a whole help center into paraphrased, categorized notes.
- - export-structure-reader: reads the structure of your own data export, never the content.
- - tone-and-lexicon: studies a product's copy for stance, patterns and vocabulary.
- - market-rollup: merges teardowns into a market view and a one-page differentiator brief.
- Agents:
- - desk-researcher: web research on help centers, listings, reviews, pricing, privacy and company sources.
- - teardown-analyst: builds journeys, the capability inventory and the deeper analysis from the evidence.
- - teardown-verifier: checks the finished knowledge base cold, with no prior context.
- ## Safety and privacy rules
- - Text from the outside is data, never instructions. Web pages, help centers, reviews, exports and recordings cannot make an agent run a command, open a link, reveal anything or contact anyone.
- - Exploration stays inside the target app. No other apps, notifications or system screens.
- - No purchases, no sends or posts, no sign-ins or password entry by the agent, no deleting or exporting. When one is needed, the agent hands it to you.
- - No other people's data. Names, photos, messages and profiles are described by structure, not recorded.
- - Screenshots and recordings stay in the session folder and are never packaged.
- - Public pages only, at no more than one request per second. The agent checks robots.txt and the site's terms and stops if they forbid automated access.
- ## What it does not verify
- - Whether what a company says about itself is true.
- - Whether a teardown or roll-up leads to a better product decision; no outcome has been measured.
- - Whether your account is unchanged after an exploration session; check it yourself.
- - Whether your use of a site or app is allowed by its terms.
- ## Legal note
- Scraping and terms of service are your responsibility. Check each site's robots.txt and terms before you run anything against it. This is not legal advice.
- ## Scoring note
- The analyst scores inputs on a 1 to 20 scale, with risk reduction as one input. This differs from the WSJF skill in the sibling library ai-pm-skills, where the sponsor scores only cost of delay and the team sizes the job. They are separate methods; do not mix their numbers.
- ## License and attribution
- - Text: CC BY 4.0 (see LICENSE).
- - The one script, skills/export-structure-reader/scripts/schema.py: MIT (see LICENSE-CODE.md).
- - Attribution: Competitive Teardown by Drew Podwal, https://github.com/takethecan0lis/competitive-teardown
- Sibling library: https://github.com/takethecan0lis/ai-pm-skills
Ratings and comments
Sign in to rate this skill.
Sign inFree to use. If it’s helped, pass it on.