Skip to content

Geo-ff/make-it-real

Repository files navigation

Make It Real

中文说明

Claude Code Codex Gemini CLI Cursor License Release LINUX DO

Turn a spark of thought into a clear path forward.

In an age where anyone can vibe code, you may have a brilliant idea you want to bring to life, but no clear sense of where to begin.

Make It Real helps you gradually uncover what it will take to make that idea real: what needs to be prepared, which technologies may be involved, what decisions matter, and what risks or constraints should be considered. You do not need to already understand product management, software architecture, or prompt design. Make It Real guides the AI assistant through choice-first questions until the whole project becomes clear enough to execute.

You bring the brainstorm. Make It Real turns it into a path.

Example Output

Make It Real English example output

Why It Exists

AI coding tools have made implementation easier, but unclear ideas still produce unclear projects. Make It Real focuses on the missing step between inspiration and execution:

  • What exactly are we making?
  • Who is it for?
  • What should the first deliverable include?
  • What should be saved for later?
  • What should a coding agent, AI assistant, or human executor do next?

What It Produces

By default, the skill creates staged Markdown documents under docs/make-it-real/:

File Purpose
00-project-blueprint.md One-page overview of the project, value, users, delivery target, and decisions.
01-product-profile.md Product/project positioning, target users, scenarios, references, constraints, and success signals.
02-scope-and-requirements.md Current delivery scope, non-goals, requirements, acceptance criteria, risks, and assumptions.
03-solution-blueprint.md Recommended implementation or execution approach, architecture, data, integrations, and operations.
04-task-and-ai-handoff.md Task list and prompts that can be handed to coding agents, AI assistants, or execution tools.

Who It Helps

  • Nontechnical founders, operators, creators, and students who have an idea but need structure.
  • Vibe coders and AI-assisted builders who want clearer tasks before opening a coding tool.
  • Product-minded developers who want a lightweight discovery workflow before implementation.
  • Agents that need a repeatable way to clarify vague project requests before writing files or code.

How It Works

  1. Ask the user a small number of choice-first questions.
  2. Adapt the depth based on the user's coding and AI-tool familiarity.
  3. Preserve the full long-term vision before narrowing the current delivery target.
  4. Define a selected delivery target such as Demo, MVP, V1, full roadmap, or another user-defined target.
  5. Summarize the plan in chat and ask for confirmation.
  6. Write final documents only after the user confirms.
  7. Explain where the documents are, what each one is for, and how to use them with AI/coding tools.

Supported Entry Points

This repository is structured as a cross-platform skill/plugin package:

.
├── .codex-plugin/
│   └── plugin.json
├── .claude-plugin/
│   ├── plugin.json
│   └── marketplace.json
├── .cursor-plugin/
│   └── plugin.json
├── AGENTS.md
├── CLAUDE.md
├── GEMINI.md
├── gemini-extension.json
└── skills/
    └── make-it-real/
        ├── SKILL.md
        ├── agents/
        │   └── openai.yaml
        ├── references/
        │   ├── ai-coding-task-template.md
        │   ├── document-templates.md
        │   └── question-flow.md
        └── scripts/
            └── create_project_docs.py

The canonical skill source is skills/make-it-real/SKILL.md.

Installation

skills CLI

Install directly with the open skills CLI:

npx skills add Geo-ff/make-it-real --skill make-it-real -y

To inspect the detected skill before installing:

npx skills add Geo-ff/make-it-real --list

This is the recommended cross-agent installation path for tools that support the open skills format. Public installs through the skills CLI also make the skill eligible for skills.sh discovery through anonymous install telemetry; the skills.sh listing may take time to refresh.

Claude Code

Install from the stable release tag:

/plugin marketplace add Geo-ff/make-it-real@v0.1.0
/plugin install make-it-real@make-it-real

For development or latest changes, omit the tag and add Geo-ff/make-it-real.

After installation, ask Claude Code to use Make It Real when clarifying a vague idea. In Claude Code environments that expose plugin namespaces directly, the skill namespace is make-it-real:make-it-real.

Gemini CLI

Install from the stable release tag:

gemini extensions install https://github.com/Geo-ff/make-it-real --ref v0.1.0

For development or latest changes, install from the default branch:

gemini extensions install https://github.com/Geo-ff/make-it-real

Codex

Use the repository as a plugin package when your Codex environment supports plugin packages. For a direct local skill install, copy skills/make-it-real/ into your Codex skills directory.

Bash example:

git clone https://github.com/Geo-ff/make-it-real.git /tmp/make-it-real
mkdir -p ~/.codex/skills
cp -R /tmp/make-it-real/skills/make-it-real ~/.codex/skills/

PowerShell example:

git clone https://github.com/Geo-ff/make-it-real.git "$env:TEMP\make-it-real"
New-Item -ItemType Directory -Force "$env:USERPROFILE\.codex\skills"
Copy-Item -Recurse -Force "$env:TEMP\make-it-real\skills\make-it-real" "$env:USERPROFILE\.codex\skills\make-it-real"

Cursor

Use .cursor-plugin/plugin.json as the plugin entry point when your Cursor setup supports plugin packages. Otherwise, reference skills/make-it-real/SKILL.md in your agent instructions.

Generic AI Agents

Use AGENTS.md as the root instruction file and point the agent to skills/make-it-real/SKILL.md.

Release

The first stable tag is v0.1.0.

Use the tag when you want reproducible installs. Use the default branch when you want the latest changes.

Usage

Example prompts:

Use $make-it-real to help me turn my rough idea for a personal finance app into a clear project plan and task handoff.
I want to build a PPT generation tool, but I only have a rough idea. Help me clarify it and create docs I can give to a coding agent.
Use Make It Real to turn this vague product idea into a confirmed V1 scope and AI-agent task list.

Document Scaffolding Script

The bundled script creates the standard project document skeleton:

python skills/make-it-real/scripts/create_project_docs.py --project-name "Project Name" --project-type "software" --language en

Use --language zh for Chinese headings. The script writes to docs/make-it-real/ by default and refuses to overwrite existing files unless --force is passed.

Validation

If you have the Codex skill creator tools installed, validate the skill folder with:

python path/to/quick_validate.py skills/make-it-real
python -m py_compile skills/make-it-real/scripts/create_project_docs.py

For Codex plugin validation, validate the repository root as a plugin package with the plugin creator tooling when available.

License

MIT

Community

Discuss Make It Real on LINUX DO.

About

A portable AI-agent skill that turns rough ideas into project specs, delivery scopes, and task handoffs.

Topics

Resources

License

Stars

Watchers

Forks

Packages

 
 
 

Contributors

Languages