agentleFS
gastonmorixegastonmorixe/project-factory/CLAUDE.md

project-factory

CLAUDE.md0 starsChanged 9 months agoMITView on GitHub

This repository is a project launcher. It exists solely to bootstrap new GitHub repositories from Claude Code Web sessions. When a user starts a Claude Code session against this repo, they will describe a new project they want to build. Your job is:

  • Commits and pushes
  • In this repositoryOne other file it gives its agents.

The file

We split this CLAUDE.md into the rules it gives an agent

5 rules. About 0 are common, 4 are distinctive, and 1 is 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.

  • 1

    Very specific

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

  • Common
  • Distinctive
  • Very specific
  1. Distinctive:

    When a user starts a Claude Code session against this repo, they will describe a new project they want to build.

    line 7
  2. Distinctive:

    Instead, use the PAT directly in the HTTPS remote URL for all operations on the new repo

    line 18
  3. Very specific:

    Use the skill at .claude/skills/create-repo/SKILL.md for the full procedure.

    line 31
  4. Distinctive:

    Never write the PAT to any file that gets committed

    line 45
  5. Distinctive:

    Never include the PAT in any git remote that gets committed to a config file

    line 46

≈ “Common” is an estimate: at least 1 in 5 of the 28,029 CLAUDE.md 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
CLAUDE.md

Project Factory

This repository is a project launcher. It exists solely to bootstrap new GitHub repositories from Claude Code Web sessions.

How This Works

When a user starts a Claude Code session against this repo, they will describe a new project they want to build. Your job is:

  1. Create a new private GitHub repo using the GITHUB_PAT environment variable
  2. Scaffold the project based on what the user described
  3. Push the scaffold to the new repo using PAT-authenticated HTTPS
  4. Tell the user to open a new Claude Code session against the new repo to continue development

Important Constraints

  • You are running inside a Claude Code Web sandbox
  • The git proxy is scoped to THIS repo (project-factory) — you CANNOT git push to other repos through it
  • Instead, use the PAT directly in the HTTPS remote URL for all operations on the new repo
  • Network access is set to "full internet" — api.github.com and github.com are reachable via curl and git
  • The GITHUB_PAT environment variable is injected via the Claude Code Web environment settings
  • NEVER echo, log, or display the PAT value

Workflow

Step 1: Validate PAT

curl -sf -H "Authorization: Bearer $GITHUB_PAT" https://api.github.com/user | jq -r '.login'

Step 2: Create the repo

Use the skill at .claude/skills/create-repo/SKILL.md for the full procedure.

Step 3: Scaffold & push

Clone the new repo with PAT auth, add files, commit, push.

Step 4: Report back

Give the user:

  • The GitHub URL
  • Instructions to start a new Claude Code Web session against it
  • A summary of what was scaffolded

Security

  • The PAT requires fine-grained permissions: Administration (read/write), Contents (read/write), and Metadata (read-only)
  • Never write the PAT to any file that gets committed
  • Never include the PAT in any git remote that gets committed to a config file
  • Clean up: after pushing, remove the cloned directory to avoid leaking the PAT in .git/config

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.