Guide · Agent skill
How to use the Setting up a project agent skill
Updated · written with AI from the project's own documentation and checked against our catalogue data
The setting-up-a-project skill is a dispatcher for JetBrains' ThinkRail workflow. When you ask an agent to set up, onboard or spec a project that has no spec graph yet, it checks what is already in the workspace and hands off to the right follow-on skill. This guide explains what the skill does, how to install it, how to trigger it, and what it needs before it will work.
Setting up a projectJetBrains/thinkrail · Apache-2.0 Download the skill Quick start
- 1Confirm the prerequisites. Make sure the ThinkRail spec tools (spec_grep and spec_graph) are available and the sibling skills are installed.
- 2Download the ZIP. Download the skill ZIP from this page.
- 3Install the skill. Unzip into ~/.claude/skills/ for personal use or into the project's .claude/skills/ folder, or upload the ZIP under Settings → Capabilities → Skills in the Claude apps.
- 4Make the request. Ask the agent in plain language to set up, onboard or spec a project that has no spec graph.
- 5Let the agent detect the situation. The agent checks for existing specs and whether the repo has real source code.
- 6Follow the handoff. The agent names one follow-on skill, or offers to review and extend existing specs.
What does the setting-up-a-project skill make the agent do?
The skill is the single entry point for turning a project into a spec graph. It does not do the setup work itself. First it detects the situation, then it follows the matching skill so the flow stays consistent.
Detection uses the ThinkRail spec tools (spec_grep and spec_graph). The agent asks two questions: is there already a spec graph (a goal-and-requirements.md or any SPEC.md), and is there real source code or is the repo empty or near-empty, such as only a README or scaffolding.
- Already has specs: the agent does not redo them. It briefly offers to review and extend the graph, for example by filling a missing architecture.md or un-specced modules. If you decline, it stops.
- No specs and an empty or near-empty repo: the agent follows starting-a-new-project, which interviews you to turn the idea into goal-and-requirements.md.
- No specs and real source code: the agent follows importing-a-codebase, which analyzes the code and agent files, draft the first spec graph, and asks only for intent the code cannot reveal.
How do you install the skill?
The skill ships as a ZIP file that you download from this page. You can install it in one of two ways, depending on whether you want it for one project or for your own agent setup.
For a personal install, unzip it into ~/.claude/skills/. To limit it to a single project, unzip it into that project's .claude/skills/ folder. Alternatively, upload the ZIP in the Claude apps under Settings → Capabilities → Skills.
- Personal: unzip into ~/.claude/skills/
- Single project: unzip into the project's .claude/skills/ folder
- Claude apps: Settings → Capabilities → Skills, then upload the ZIP
How do you trigger the skill with a natural request?
You do not need a special command. Ask the agent in plain language to set up, onboard, initialize or spec a project that has no spec graph. The skill's description covers these phrasings, so requests like the ones below should route to it.
The skill is also invoked when the app's Set-up-project card is used. It is not meant for ordinary work in a project that already has specs, and the skill itself says not to use it there.
- "Set up this project for ThinkRail."
- "Onboard this existing repo into the spec-driven workflow."
- "Initialize a spec graph for this project."
- "Spec out this new project before we write code."
What do you need before using it?
The skill depends on other pieces that are not included in the provided files. Check these before you rely on it.
The skill requires the ThinkRail spec tools (spec_grep and spec_graph). It also routes to four sibling skills: starting-a-new-project, importing-a-codebase, brainstorming and writing-specs. Those sibling skills are not in this file, so the dispatcher will not complete a handoff without them.
- ThinkRail spec tools: spec_grep and spec_graph
- Sibling skills: starting-a-new-project, importing-a-codebase, brainstorming, writing-specs
- Projects inside the ThinkRail app ecosystem; the skill is not designed for projects outside it
What are the safety notes?
This skill runs no scripts and makes no network calls, according to its listing. It works by reading the workspace and using the spec tools, then handing off to other skills.
Because it reads your code and agent files to draft a spec graph, review the drafts the agent produces before you accept them, especially in a repo you care about.
- Runs scripts: no
- Uses network: no
- It reads workspace files and the spec tools to detect the situation
What tips help the setup go smoothly?
Run the skill once per project, at the start. If a spec graph already exists, expect an offer to extend it rather than a full rebuild.
Answer the interview questions with intent the code cannot show, such as goals and requirements. The importing path is designed to ask only for that kind of information.
- Use it at the start of a project, not in the middle of feature work.
- If specs exist, accept the review offer to fill gaps, or decline to stop.
- Keep the sibling skills installed alongside this one.
What people use it for
- Onboarding an existing repo with real code into a ThinkRail spec graph via importing-a-codebase.
- Turning a new, near-empty repo idea into goal-and-requirements.md via starting-a-new-project.
- Adding a missing architecture.md to a project that already has some specs.
- Giving a team one consistent way to create its first spec graph.
Frequently asked questions
Is the setting-up-a-project skill free?
The source lists the licence as Apache-2.0 but does not state any price. Check the repository's licence file for the terms that apply to you.
What do I need to run this skill?
You need the ThinkRail spec tools (spec_grep and spec_graph) and the sibling skills starting-a-new-project, importing-a-codebase, brainstorming and writing-specs. The sibling skills are not included in the files provided here.
Where do I install the skill?
Unzip the downloaded ZIP into ~/.claude/skills/ for personal use, or into a project's .claude/skills/ folder. You can also upload it in the Claude apps under Settings → Capabilities → Skills.
Does this skill work outside ThinkRail?
Not as listed. It is built for projects in the ThinkRail app ecosystem and depends on ThinkRail spec tools and sibling skills that are not included in this file.
Why did the agent offer to review my specs instead of setting up the project?
The skill detects an existing spec graph, such as goal-and-requirements.md or a SPEC.md, and does not redo it. It offers to fill gaps like a missing architecture.md. If you decline, it stops.
Should I use this skill for a bug fix or feature in my project?
No. The skill is for setting up or onboarding a project with no spec graph. Ordinary work in an already-specced project should not use it, and the skill says so. Feature and behaviour work routes through brainstorming.
New agent skills, repos and AI tools — every week
The best new skills and open-source projects, the fastest-growing repos, and one free prompt. Five minutes to read. Unsubscribe anytime.
Our verdict
Use this skill if you are adopting ThinkRail and want a consistent first step for creating a spec graph, either in an empty repo or an existing codebase. Pick something else if your project already has specs and you just need ordinary feature work, or if you are not working in the ThinkRail ecosystem, since the required sibling skills and spec tools are not included here.
Download the skill