agentsclimarketplace

Project init

Skill alfredang/skills/project-init

Claude Code skills: auto-detect & start any app, README generator, Docker Hub push, LinkedIn posts, NotebookLM slides, secret scanner, mobile responsiveness testing | Install with npx skills add

Install
npx -y skills add alfredang/skills --skill project-init

Assembled from the repository path, not quoted from the project. Check it against their README if it does not work.

2 things to look at

  • no licenseNo license file was found in the repository. Code published without one is not open source by default, so using it at work is a question for whoever answers licensing questions where you are.
  • 2 stars2 stars. Stars are a popularity signal and not a quality one, but at this level it is likely that nobody has read this closely except its author, and you would be relying on your own review.

What its author says it does

Copied from the file, not written here

Initialize a new project with standard CLAUDE.md, MEMORY.md, and .env setup. Use when starting a new project or setting up memory and secrets management for an existing project.

SKILL.md

3.3 KB, as published. Nobody here has run it

Project Init Skill

Set up the standard project scaffolding for memory management, instructions, and secrets. Run the following steps:

Step 1: Detect Project Info

  • Determine the project name from $ARGUMENTS if provided, otherwise use the current directory name.
  • Determine the absolute path of the current working directory.
  • Compute the Claude memory path: ~/.claude/projects/<sanitized-project-path>/memory/ where <sanitized-project-path> is the absolute path with / replaced by - and leading - preserved (e.g., /Users/alice/projects/my-app becomes -Users-alice-projects-my-app).

Step 2: Create CLAUDE.md

Create CLAUDE.md in the project root with the following content (replace <MEMORY_PATH> with the computed memory path):

# Project Instructions

## Memory Management
- Always read `<MEMORY_PATH>/MEMORY.md` at the start of each conversation for context from previous sessions.
- Update MEMORY.md with key decisions, preferences, and insights discovered during the session.
- Keep MEMORY.md under 200 lines — trim outdated or redundant entries to stay within this limit.
- For detailed notes, create separate topic files in the memory directory (e.g., `debugging.md`, `architecture.md`) and link from MEMORY.md.
- Do not duplicate information already present in this CLAUDE.md file into MEMORY.md.
- Before adding new memories, check if an existing entry can be updated instead.
- Remove memories that are no longer accurate or relevant.

## Secrets Management
- All API keys and secrets go in `.env` (already gitignored).
- Never hardcode secrets in source code or commit `.env` to git.
- Reference environment variables in code instead of raw values.

If CLAUDE.md already exists, merge the Memory Management and Secrets Management sections into it without overwriting existing content.

Step 3: Create MEMORY.md

Create the memory directory at the computed path, then create MEMORY.md inside it:

# Project Memory

## Project Overview
- **Project**: <project-name>
- **Location**: <absolute-path>

## User Preferences
<!-- Add user preferences here -->

## Key Files
- `CLAUDE.md` — project-level instructions loaded every session

## Architecture & Decisions
<!-- Add key architectural decisions here -->

## Topic Files
<!-- Link to detailed topic files as needed -->
<!-- Example: - [debugging.md](debugging.md) — recurring issues and fixes -->

If MEMORY.md already exists, do not overwrite it.

Step 4: Create .env

Create a .env file in the project root with placeholder content:

# Project Secrets — DO NOT COMMIT
# Add your API keys and secrets below

If .env already exists, do not overwrite it.

Step 5: Update .gitignore

Ensure .gitignore exists and contains these entries (add only missing ones):

.env
.DS_Store

Step 6: Summary

Report what was created or already existed:

  • CLAUDE.md (created / already existed — merged)
  • MEMORY.md at <MEMORY_PATH> (created / already existed — skipped)
  • .env (created / already existed — skipped)
  • .gitignore (updated / already had entries)

Keep looking

Skills are one crate of 328,083. Ordering is by how many stacks a row turns up in, so the top of any crate is what has actually been picked rather than what has the most stars.