agentleFS
srikanth-manivannansrikanth-manivannan/claude-agent-factory/skills/developer-experience/project-scaffolder/SKILL.md

project-scaffolder

Skill1 starsChanged 5 months agoMITView on GitHub

Use this to bootstrap a new project from a stack profile with structure, lint, CI, and docs baked in.

  • Reads credentials

6 other projects use a version of this.

Adopted by

Most recently taken up by

The file

Files
SKILL.md

Project Scaffolder

Use this to bootstrap a new project from a stack profile with structure, lint, CI, and docs baked in.

Tech profile — Technology: project scaffold · Language: any · Stack: any · Toolchain: any · Domain: developer-experience (Stack-agnostic method; emits a layout + config for the chosen stack via the tech profile.)

When to use this skill

  • Starting a new service/library/app and wanting consistent, golden-path conventions.
  • Standardizing how an org bootstraps repos (structure, lint, CI, docs, conventions).
  • Producing the baseline the scaffold-new-service workflow builds on.

When NOT to use this skill

  • Configuring CI for an existing repo — use devops/ci-pipeline-generator.
  • Setting up only lint/format — use developer-experience/linter-formatter-setup (planned).
  • Provisioning infra — use cloud/cloud-resource-provisioner.

Prerequisites

  • The project type (service/library/CLI/app) and target stack.
  • Any org conventions to enforce (license, code owners, commit style).

Inputs

InputRequiredDescription
project_typeyesservice / library / CLI / app
stackyesLanguage/framework/toolchain
nameyesProject name (kebab-case)
conventionsnoOrg rules: license, CODEOWNERS, commit/branch policy

Instructions

  1. Choose the golden-path layout for the project type + stack: source, tests, config, and docs directories with predictable names.
  2. Add a README + docs skeleton following /standards/documentation.md (what/quickstart/inputs-outputs/limitations) — the project is documented from commit one.
  3. Wire lint + format + pre-commit for the stack so style is automatic, not argued.
  4. Add a test harness with one passing example test (so test works day one) — use testing/unit-test-generator for the example.
  5. Add CI via devops/ci-pipeline-generator (install → lint → test → build), set as a required check.
  6. Bake in conventions: LICENSE, .gitignore, editorconfig, CODEOWNERS, and a CHANGELOG — plus commit/branch policy if provided.
  7. Add a task runner (make/just/npm-scripts) exposing setup, test, lint, build, run so every project has the same verbs.
  8. Verify the golden path: a fresh clone can setup then test green with no tribal knowledge.

Output

A ready-to-commit project scaffold: golden-path directory layout, README + docs skeleton, lint/format/pre-commit config, a test harness with a passing example, CI config, baked-in conventions (LICENSE/.gitignore/CODEOWNERS/CHANGELOG), and a task runner with standard verbs. A fresh clone is productive immediately.

Constraints & safety

  • Golden path works on a fresh clone — setup → test green, no hidden steps.
  • Conventions are enforced, not suggested (lint/CI/pre-commit), so quality is automatic.
  • No secrets in the scaffold; provide .env.example, never .env (/standards/security.md).

Examples

Minimal below; full enterprise examples in EXAMPLES.md; build steps in IMPLEMENTATION.md.

Given: project_type = service, stack = "<your backend stack>", name = billing-svc. Produces: src/, tests/ (with a passing example), lint+pre-commit, CI (lint→test→build, required), README/docs, LICENSE/.gitignore/CODEOWNERS, and a task runner — setup && test green on a fresh clone.

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.