Answer 8 questions and get a personalized tool recommendation, risk profile, and setup checklist. No technical knowledge required.
Question 1 of 80%
Question 1 of 8
How comfortable are you with software setup and troubleshooting?
Be honest — this shapes which setup will actually work for you long-term.
Question 2 of 8
Do you have a technical person available to help set this up and maintain it?
Having someone technical nearby changes the risk picture significantly — they can configure things safely that most people can't do solo.
Question 3 of 8
Should the agent start fresh, or access your existing files?
This is the single biggest factor in how much damage a worst-case scenario can cause.
Question 4 of 8
How often will the agent read untrusted content — web pages, emails, or files from strangers?
Prompt injection attacks — where a malicious file tricks the agent into doing something harmful — are a real, documented threat for both tools.
Question 5 of 8
How do you feel about an AI company's servers processing the work you send the agent?
Cowork always sends your work to Anthropic's cloud. OpenClaw can run entirely on your own machine — but only if set up that way by someone technical.
Question 6 of 8
What's your worst-case nightmare if something went wrong?
Pick the one that would genuinely hurt you the most.
Question 7 of 8
Will you connect this assistant to the internet, chat apps, or shared networks?
OpenClaw has a "gateway" that, when misconfigured or publicly exposed, has thousands of documented compromised instances in the wild.
Question 8 of 8
How much control over your computer does the agent actually need?
More power means a bigger blast radius if something goes wrong. Match the tool to what you actually need, not what sounds impressive.
Your Results
Fit Scores
Cowork
—
OpenClaw
—
Risk Level
—
Watch These Risks
Sharing this quiz: Download the HTML file and send it to anyone — it runs entirely in the browser with no tracking and no data sent anywhere. To convert it to a Google Form, copy each question and add a final section that shows the result text based on the score.