agentleFS
T4LELT4LEL/Claude-Arsenal/skills/new-project/SKILL.md

new-project

Skill1 starsChanged 4 months agoMITView on GitHub

Use when starting any new app, webapp, API, tool, or side project from scratch - when the user names an idea to build or says new project, scaffold, start a repo, set up a project. Produces a running, git-tracked, Claude-ready project skeleton.

  • Reads credentials
  • Installs packages

51 other projects use a version of this.

Adopted by

Most recently taken up by

Used with

Projects with it also use checkpoint, prd and 10 more.

Other files people use alongside this one.

Files found in the same public projects as a version of this one, more often than they turn up anywhere else. Each project counts once, large collections count for less, and this file's own repository is left out.
  • genomewalkerSkill

    checkpoint

    Save a checkpoint of task status, decisions, active files, and next steps

    Also in 3 projects that use this

  • GitHubSkill

    prd

    Generate high-quality Product Requirements Documents (PRDs) for software systems and AI-powered features. Includes executive summaries, user stories, technical specifications, and risk analysis.

    Also in 3 projects that use this

  • sunnyclinches44Skill

    explain

    Explains code using real-world analogies and ASCII diagrams

    Also in 3 projects that use this

  • SentrySkill

    spec

    Produce a verified spec — problem, desired outcome, and acceptance criteria — before building. Use when given a GitHub/Linear issue link, when starting from a fuzzy in-conversation idea, or asked to "figure out what to do" / "scope this" / "what should I build" — whenever the *what* isn't yet pinned down. Grills the ambiguities, checks the premise against the code, and produces acceptance criteria you sign off on — then hands to design-first.

    Also in 3 projects that use this

  • GoogleSkill

    domain-modeling

    Build and sharpen a project's domain model. Use when the user wants to pin down domain terminology or a ubiquitous language, record an architectural decision, or when another skill needs to maintain the domain model.

    Also in 4 projects that use this

  • SentrySkill

    review

    Fixture review skill.

    Also in 5 projects that use this

  • +1 more from getsentry

The file

We split this skill into the rules it gives an agent

6 rules. 0 are common, 2 are distinctive, and 4 are very specific to this project.

  • 0

    Common

    Most of the 48 edited versions we read keep these.

  • 2

    Distinctive

    This author's own choices. Worth a look.

  • 4

    Very specific

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

  • Common
  • Distinctive
  • Very specific
  1. Distinctive:

    Use what the user already provided (including arguments to this skill).

    line 26
  2. Very specific:

    Paid product?: If yes, note it — Step 6 points at /monetize.

    line 31
  3. Very specific:

    If the product hasn't been brainstormed yet in this conversation, invoke superpowers:brainstorming for the product itself BEFORE …

    line 35
  4. Very specific:

    Check your work: search the new CLAUDE.md for {{ — it must return zero matches.

    line 66
  5. Distinctive:

    Do NOT commit — per global rules the user commits explicitly.

    line 68
  6. Very specific:

    Run the real check and paste its real output.

    line 72

“Common” is measured: kept by at least 6 in 10 of the 48 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

New Project

Take a raw idea to a running, git-tracked, Claude-ready project skeleton.

Templates are bundled with this skill in templates/ (synced copies — the source of truth is the toolkit repo at X:\Projects\ClaudeCode\templates\; edit there, then re-run sync.ps1).

Copy this checklist and check off items as you complete them:

New Project Progress:
- [ ] Step 1: Gather requirements
- [ ] Step 2: Design gate
- [ ] Step 3: Scaffold by type
- [ ] Step 4: CLAUDE.md + git init
- [ ] Step 5: Verify it runs (mandatory)
- [ ] Step 6: Hand off

Step 1 — Gather

Use what the user already provided (including arguments to this skill); ask only for what's missing.

  • Name (kebab-case) and one-line purpose.
  • Type: webapp (default), api, or other.
  • Location: default X:\Projects\<name>.
  • Paid product? If yes, note it — Step 6 points at /monetize.

Step 2 — Design gate

If the product hasn't been brainstormed yet in this conversation, invoke superpowers:brainstorming for the product itself BEFORE scaffolding. Never scaffold an undesigned product. If a design/spec already exists, proceed.

Step 3 — Scaffold by type

If a scaffold command fails, stop and fix it before continuing: read the exact error; for generator flag errors run the generator's --help and adapt — flags drift between versions and the intent below matters more than the exact flags. Never leave a half-scaffolded folder silently: finish it, or delete it and report why.

webapp (Next.js + TypeScript + Tailwind + Supabase):

npx create-next-app@latest <name> --typescript --tailwind --eslint --app --src-dir --import-alias "@/*" --use-npm --yes
cd <name>
npm install @supabase/supabase-js @supabase/ssr

Then create .env.example (NEXT_PUBLIC_SUPABASE_URL, NEXT_PUBLIC_SUPABASE_ANON_KEY, commented), src/lib/supabase/ client helpers per current @supabase/ssr docs (fetch them via the context7 MCP tools — the cookie/session patterns change), and a supabase/migrations/ folder. Only create the actual Supabase project when the design needs a database — ask before creating cloud resources.

api (Python + FastAPI):

mkdir <name>; cd <name>
python -m venv .venv; .\.venv\Scripts\Activate.ps1
pip install fastapi "uvicorn[standard]" pytest httpx; pip freeze > requirements.txt

Structure: app/main.py (FastAPI instance + health route), app/routers/, app/models.py, tests/test_health.py, .env.example (commented var names, no values). Add app/db.py and a migrations setup only when the design includes a database.

other (CLI, extension, desktop, mobile): consult templates/project-structures.md for the layout, scaffold minimally with the ecosystem's official generator.

Step 4 — Claude-ready

  1. Copy the matching template from this skill's templates/ — CLAUDE.webapp.md, CLAUDE.api.md, or CLAUDE.generic.md — to the project root as CLAUDE.md and fill every {{placeholder}}.
  2. Check your work: search the new CLAUDE.md for {{ — it must return zero matches.
  3. git init; ensure .gitignore covers .env* (except .env.example), node_modules/.venv, build output.
  4. Do NOT commit — per global rules the user commits explicitly.

Step 5 — Verify (mandatory)

Run the real check and paste its real output. webapp: npm run build. api: pytest. other: the ecosystem's build/test command.

  • Passes → continue to Step 6.
  • Fails → fix and re-run until it passes. If you cannot make it pass, report the exact failing output and what you tried — never hand off a skeleton claimed as working without passing output.

Step 6 — Hand off

Report: location, what was scaffolded, verification output. Suggest next actions: run /product-spec for anything non-trivial, write the first vertical slice (superpowers:writing-plans), and if it's a paid product, run /monetize BEFORE building features — pricing shapes scope.

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.