agentleFS
wald21x6-opswald21x6-ops/builder-kit/skills/app/SKILL.md

app

Skill1 starsChanged 23 days agoMITView on GitHub

Playbook for web apps and dashboards: CRMs, booking or ordering systems, customer portals, admin panels, owner dashboards, internal tools, phone-style web apps (PWAs), directories and marketplaces, 3D product configurators. Adds roles-and-permissions planning, a backend preflight, testing every kind of user, "built / deployed / live-verified" status per feature, and dashboards designed around decisions. Use it with the build skill whenever the user wants software people log into or use repeatedly, even if they call it a "website with a dashboard" or "a system for my business".

4 other projects use a version of this.

Adopted by

Most recently taken up by

  • Other versions4 projects use a version of this.

    Word for word · 0

    None found.

    Edited · 0

    None found.

    Wrote their own version · 4

    1. nmamano/isomux★ 43same name
    2. aranculleton/last-job-search★ 6same name
    3. AxlZero21/Skills★ 1same name
    4. GustavoAlecio/claude-setup★ 0same name

    How matches are found. Word for word: the exact same text. Edited: shares passages with this file (“99% same” means 99% of its five-word phrases also appear here; 10% or more counts). Their own version: similar meaning, or, weakest, the same name. We can’t tell which project had it first. Projects are counted among the 200 most-starred candidates, and mentions are links to the repository we found, up to 50 per site, so both counts are a floor.

  • In this repository7 other files it gives its agents.

The file

We split this skill into the rules it gives an agent

6 rules. 0 are common, 5 are distinctive, and 1 is very specific to this project.

  • 0

    Common

    Most of the 3 edited versions we read keep these.

  • 5

    Distinctive

    This author's own choices. Worth a look.

  • 1

    Very specific

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

  • Common
  • Distinctive
  • Very specific
  1. Very specific:

    Use this with build, and load backend-login for anything involving sign-in, data storage, payments or going live.

    line 9
  2. Distinctive:

    Write a short data plan in plain words (what's stored, who can read it, who can change it) and a user-states list.

    line 20
  3. Distinctive:

    Keep a feature status table in CURRENT.md with three separate columns: built · deployed · live-verified.

    line 29
  4. Distinctive:

    Use an independent critic on the images, and agree a stop rule with the user ("good enough for now, move on").

    line 37
  5. Distinctive:

    every number must change what someone does next

    line 41
  6. Distinctive:

    Clean up after testing: no locked accounts, test emails or junk rows left behind, and say what you created.

    line 50

“Common” is measured: kept by at least 6 in 10 of the 3 edited versions of this file we could read. “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

App and dashboard playbook

Use this with build, and load backend-login for anything involving sign-in, data storage, payments or going live. <kit> = the Builder Kit folder named in the session-start notice (usually ~/.claude/skills/builder-kit).

Target (extra questions)

  • Who uses it, and what can each kind of user see and do? List the roles: visitor, customer, staff, owner.
  • Where does the data come from today? WhatsApp, a spreadsheet, the owner's head. Moving it is often half the job.
  • For a dashboard: which decision does each screen help someone make? "Just an inquiry dashboard" got rejected because the owner wanted a dashboard for the whole business. The real goal surfaced only after 3h40 of questions.
  • First real use: the first real customer order tracked, the first real lead logged, the first staff login.

Plan (extra)

  • Write a short data plan in plain words (what's stored, who can read it, who can change it) and a user-states list. Both go on the Task Card.
  • Riskiest part first: usually the data path (a lead form arriving in the database) or the login. Build and live-test it before any polish. In the best build in the data, the lead pipeline was proven on day one.
  • Stack default: Next.js or Astro front end, Supabase for login and database, hosted on Vercel or Cloudflare. Simpler is better: a spreadsheet plus a form beats an app nobody finishes. Say what you chose and why in one line.
  • Run the backend preflight before the first feature: references/backend-preflight.md.

Build (extra)

  • Keep a feature status table in CURRENT.md with three separate columns: built · deployed · live-verified. "Verified with test stubs" isn't verified. In one app, every brief passed against stubs while production was broken for four days.
  • No green indicator unless the real path produced it. One key check showed green for keys that then failed, because it only tested listing models, not actually generating.
  • Several AI tools on one codebase? Give each tool a folder or branch in CURRENT.md, check git status for changes that aren't yours, and deploy only what you made and checked.
  • Visual-heavy apps (3D, image tools): "done" is a sentence a stranger could test. "10 out of 10" isn't one. Pair automated checks with looking: 45 of 45 checks passed while black specks and inside-out pockets were visible. Use an independent critic on the images, and agree a stop rule with the user ("good enough for now, move on").

Dashboards

Follow references/dashboards.md:

  • every number must change what someone does next
  • show the Home screen first and get a look before building the rest
  • label sample data as sample data
  • for owners who've never used a dashboard, make it one tap per task

Prove it (app extras)

  • Test every user state: visitor, new sign-up, returning user, owner/admin, wrong account, signed out. Clear the session between them. A "button missing" bug only showed up for an owner who was already signed in.
  • One real request through production for each critical flow (sign up, save, the owner sees it), using real keys.
  • Clean up after testing: no locked accounts, test emails or junk rows left behind, and say what you created. Five deliberate wrong passcode tries once locked the live dashboard just before the user tried it.
  • Re-run the backend preflight before calling it done.

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.