Skill Community skill

Jakeschincariol/replica-skill

Eleven free, MIT-licensed Claude skills that reverse-engineer an app, rebuild it clean-room, test it, score parity, mine user complaints, rebrand and deploy.

  • 1.4k GitHub stars
  • Python
View on GitHub (opens in a new tab) Setup guide
Jakeschincariol/replica-skill preview image

What it is

Replica is a set of eleven Claude skills that take an app from reverse-engineering through rebuild, testing, rebranding, launch and deployment. Each skill reads what the previous one wrote into a replica/ folder in your project. It rebuilds features and flows clean-room style, not the original's code, logo, copy or content. One skill, the Entrepreneur, reads real user reviews and turns the complaints into fixes and a positioning angle.

Who it's for

  • Developers or founders who want to rebuild an existing app's features and flows as their own product
  • People evaluating whether an app is worth cloning, using /replica-entrepreneur on its own
  • Claude Code users who want a staged pipeline from recon to deployment

Requirements

Requirements

  • Claude (Claude Code for plugin or skills-folder install; without it you can paste a single SKILL.md at the top of a chat)
  • Python 3.8 or newer for the tools; nothing to pip install

Setup

  1. Install by pasting into Claude

    Paste this into Claude.

    bash
    https://github.com/Jakeschincariol/replica-skill

install skill

  2. Install as a Claude Code plugin

    Skills are namespaced, e.g. /replica-skill:replica-recon.

    bash
    /plugin marketplace add Jakeschincariol/replica-skill
/plugin install replica-skill@replica-skill

  3. Copy folders for plain command names

    Gives plain /replica-recon style commands. For project-local use, copy the same folders into your repo's .claude/skills/.

    bash
    git clone https://github.com/Jakeschincariol/replica-skill.git
cp -r replica-skill/replica-* ~/.claude/skills/

Examples

Pipeline order

Prompt
prompt
recon -> architect -> design -> build -> backend -> test -> diff -> entrepreneur -> brand -> launch -> deploy

Expected output: Run the skills in this order; each reads what the previous one wrote in the replica/ folder.

Score parity against the original

bash
bash
python3 replica-diff/parity.py replica/features.csv                     # parity score + missing list

What it does: Weights must, should and could features, counts partial as half, and flags missing must-haves as not shippable.

Rank what users hate

bash
bash
python3 replica-entrepreneur/reviews.py replica/reviews.csv             # what users hate, ranked, linked

What it does: Drops reviews without a link and prints quotes copied from your rows with their URLs; themes with under three reviews or one source are marked thin.

Sweep for leftovers of the original

bash
bash
python3 replica-brand/sweep.py . --avoid "Original App"                 # anything of the original left?

What it does: Finds the original's name, domain and colours in your code, including inside identifiers, and blocks deploy until clean.

Pros & cons

Pros

  • Pro:Covers the full path from recon to deployment, with each skill reading the previous one's output
  • Pro:Includes standard-library Python tools (image diff, parity score, review ranking, brand sweep, listing check) that do not touch the network
  • Pro:Built-in guardrails: clean-room rebuild, public sources only, mandatory rebrand, and deploy blocked until the sweep is clean
  • Pro:Free, MIT, no signup or API key

Cons

  • Con:Does not clone content, networks or licences an app owns, such as a music app's catalogue
  • Con:Results depend on app complexity; a booking tool takes weeks and a spreadsheet engine is far harder
  • Con:Not legal advice; trademark checks and a lawyer are still needed before selling

Images

Jakeschincariol/replica-skill preview image
Jakeschincariol/replica-skill preview image · Jakeschincariol/replica-skill on GitHub
Jakeschincariol/replica-skill preview image