agentleFS
dmminga1983dmminga1983/minga-replica/skills/replica-recon/SKILL.md

replica-recon

Skill1 starsChanged 18 days agoMITView on GitHub

Reverse-engineer any app into a recon map: screen inventory, user flows, components, inferred data model and a feature matrix, from public pages, store listings, help docs and the user's own account, plus the region it will be sold in. The first step of minga-replica. Use for 'clone this app', 'reverse engineer X', 'map out X', 'build my own version of X'; RU: клон приложения, разбери приложение, сделать аналог, как устроено приложение.

2 other projects use a version of this, 2 of them edited.

Adopted by

Most recently taken up by

The file

We split this skill into the rules it gives an agent

4 rules. About 0 are common, 4 are distinctive, and 0 are very specific to this project.

  • ≈0

    Common · estimate

    Many files like this have a rule of this kind.

  • ≈4

    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:

    Public sources and the user's own account only: Never another person's account, never a password, never past a paywall or a login by any …

    line 24
  2. Distinctive:

    Reading, not scraping: No crawlers, bulk downloads or loops.

    line 26
  3. Distinctive:

    No source code, no private APIs: Do not save or read the app's JavaScript bundles, decompile its binary or log its network calls to copy …

    line 28
  4. Distinctive:

    Check the terms: If the user's account terms forbid using it to build a competitor, say so and work from public sources only.

    line 30

≈ “Common” is an estimate: at least 1 in 5 of the 350,495 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

replica-recon

Everything else in the pack builds on what this writes. A bad recon map means a bad clone.

Output in the user's project:

replica/region.txt      ru, global or both
replica/recon.md        the recon map (template: recon-map.md in this folder)
replica/features.csv    the feature matrix (template: features.csv in this folder)
replica/screens/        reference screenshots of the original. Never shipped.

The rules, before anything else

This rebuilds functionality and UX patterns, clean-room style: what the app does and how a user moves through it. Never what the app owns.

  • Public sources and the user's own account only. Never another person's account, never a password, never past a paywall or a login by any trick.
  • Reading, not scraping. No crawlers, bulk downloads or loops. With a browser tool, read at human speed with the user present.
  • No source code, no private APIs. Do not save or read the app's JavaScript bundles, decompile its binary or log its network calls to copy endpoints. Public API docs are fine.
  • Check the terms. If the user's account terms forbid using it to build a competitor, say so and work from public sources only.
  • Screenshots are reference: for layout comparison, never in the clone.

Step 1: scope, three questions

  1. Which app and platform: web, iOS, Android.
  2. Which slice. "All of Notion" is not a project; "pages, blocks and sharing" is. Default to the core loop: the one flow users pay for.
  3. Where it will be sold: Russia (ru), the rest of the world (global), or both. Write replica/region.txt. Selling in Russia changes the stack (personal data stored in Russia), the payments (54-FZ receipts) and the stores (RuStore; Google Play does not pay Russian accounts). Decide it now, not at deploy.

Step 2: sources table, a URL on every row

Help center and docs (the fullest feature list there is), pricing page (which features matter: they gate them), changelog, store listings (App Store, Google Play, RuStore), public walkthrough videos, marketing site, public API docs, the user's own account.

Step 3 onwards

  • Screens S01, S02... with route, purpose, components and every state seen: empty, loading, filled, error, no permission, mobile. An empty state not recorded is an empty state not built.
  • Flows F01... as screen sequences with edge cases. Count the clicks on the happy path: the number to beat.
  • Components: name, variants, states, which screens use it.
  • Inferred data model: entities and fields, each with its evidence and a confidence. Guesses are marked as guesses.
  • Feature matrix features.csv: feature, area, priority (must/should/could), original, clone (starts at no), notes.
  • Out of scope, honestly, as skip rows with a reason: licensed content, the network and its users, data the app owns, partner deals, regulated licences.
  • Size: S a weekend, M weeks, L a quarter, XL rescope. No promise of a perfect clone.

Output

The files above and five lines: the core loop, screen and flow counts, the three hardest parts, what is out of scope, the region. Next: /replica-architect.

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.