How to use the Productize Yourself agent skill

What does the productize-yourself skill make the agent do?

The skill switches the agent into a coaching framework when you are unsure which direction to pursue or want a niche that is hard to replace. Its core formula is special knowledge multiplied by productization (leverage such as code, media, or products) produces a hard-to-replace position.

It asks the agent to help you find activities that feel like play to you but look like work to others, check whether they are hard to train, and test whether they can be scaled. It then suggests publishing a small piece of work within 30 days and revisiting the direction every quarter.

• Identify special knowledge: things that cannot be trained and that feel like play to you
• Check uniqueness: are you one of very few people who can do this?
• Check scalability: can code, media, or a product amplify it?
• Publish a minimal piece of work within 30 days, under your own name
• Re-ask every quarter whether you are doing it as well as anyone in the world

How do you install the productize-yourself skill?

Download the ZIP file from the item page on this directory and unzip it. The package contains SKILL.md, test-prompts.json and test-results.md.

For a personal install, place the unzipped folder in ~/.claude/skills/. For a single project, place it in that project's .claude/skills/ folder. Alternatively, in the Claude apps, open Settings, then Capabilities, then Skills, and upload the skill there.

• Personal: ~/.claude/skills/productize-yourself
• Project only: .claude/skills/productize-yourself inside your project

How do you trigger the skill with natural requests?

You do not need a special command. Describe your situation in plain language and the agent should recognise it. The skill's trigger words include 找方向, 独特优势, 副业, 不可替代, productize, special knowledge and moat, and it also responds to requests like the ones below.

Good prompts state the uncertainty directly. Mention what you currently do and what you are choosing between, so the agent has material to work with.

• "Help me find my special knowledge. I am a designer and I don't know whether to freelance or build a product."
• "I don't want to keep doing work that can be replaced. How do I build a distinctive niche?"
• "What content or small product should I launch? Find what others can't easily copy."
• "How do I find my moat? I'm choosing a side hustle."

What should you expect from a session?

The agent works through the four steps in the skill: inventory your special-knowledge signals, test uniqueness against scalability, define a minimal product to publish within 30 days, and revisit the definition of your work each quarter. Expect questions back, since the framework depends on your own answers and on asking people around you what they think you are good at.

The skill is coaching, not a data-driven career test. Its output is a direction to test, not a verdict.

What does the skill need, and is it safe to run?

The skill lists no requirements. It does not run scripts and does not use the network, according to its listing. It is a set of instructions in SKILL.md, so there is little to review beyond reading that file.

It is released under the MIT licence.

• Requires: nothing listed
• Runs scripts: no
• Uses network: no

When should you not use this skill?

Do not use it if you already have a clear direction and only need execution details. The skill itself says the agent should give execution advice in that case. It also excludes job applications, resumes and interview preparation.

The framework also has limits. It assumes capital and early-access advantages that not everyone has. Finding special knowledge does not guarantee that the market will pay for it.

• Clear direction already: ask for execution help instead
• Job search, resumes, interview prep: out of scope
• Expecting guaranteed income: not what this skill offers

Tips for getting useful results

Bring concrete examples rather than abstract goals. The skill is built around what feels like play to you and looks like work to others, so describe specific activities. Ask people who know you what they think you are good at, because the framework itself suggests this cross-check.

Treat the 30-day publishing step as the real test. A small public piece of work will tell you more than another round of thinking.

• Name three to five specific activities you enjoy
• Check your answers with two or three people who know you well
• Keep the first published piece small and signed with your name