SCW's MCP Server Prompt Library

This article gives you ready-to-use prompts for the SCW MCP Server, grouped by the question you want answered. Copy a prompt into your AI tool, replace anything in [square brackets], and send it. Each prompt explains why it's worded the way it is, so you can write your own.

Here's what it looks like when you send one of these prompts:

In this article:

1. How active are my developers?
2. How are my Quests going?
3. Where can we improve?
4. Tournaments, Assessments and Cybermon
5. Commit and vulnerability activity
6. AI usage reporting
7. Tips for better answers

How active are my developers?

Using the SCW connector, how engaged were our learners in the last 90 days?

Why it works: A 90-day window is long enough to show a real pattern and short enough to act on.

Using the SCW connector, how many enabled learners haven't been active in 
the last 60 days? Group them by team.

Why it works: Asking for enabled learners leaves out disabled accounts, and grouping by team tells you which managers to contact.

How are my Quests going?

Using the SCW connector, show progress on our [Quest name] by team, and 
list anyone who hasn't started.

Why it works: Naming one Quest keeps the answer focused, and the list of people who haven't started is the list you'll follow up with.

Using the SCW connector, give me a short training summary for [team name] 
that I can use in my 1:1s this week.

Why it works: Naming the team and the purpose tells the AI tool how much detail you need and how to present it.

Using the SCW connector, draft a PCI DSS Quest for our Java developers. Save 
it as a draft so I can review it in SCW.

Why it works: Naming the standard and the language means the Quest covers the right topics in the right language. Asking for a draft keeps you in control of what gets published.

Using the SCW connector, we see too many CWE-79, CWE-89 and CWE-352 findings 
in our code. Draft a Quest that covers them and save it as a draft.

Why it works: Naming the CWEs tells the AI tool exactly which vulnerabilities to cover, and asking for a draft keeps you in control of what gets published.

Where can we improve?

Using the SCW connector, show Trust Score by team, from highest to lowest.

Why it works: Sorting the teams shows at a glance where to focus.

Using the SCW connector, which vulnerability categories have the lowest 
accuracy across our developers? Suggest training for the bottom 3.

Why it works: Accuracy shows real knowledge gaps, and limiting it to 3 categories gives you a short, practical training plan.

Using the SCW connector, what is our Trust Score, how does it compare with 
our industry, and what are the top 3 ways to improve it?

Why it works: Adding the industry comparison puts your score in context, and asking for 3 actions turns it into a plan.

Tournaments, Assessments and Cybermon

Using the SCW connector, compare our last 3 tournaments by vulnerability 
category accuracy. What improved and what got worse?

Why it works: Comparing several tournaments shows trends that one event can't.

Using the SCW connector, show pass rates for our latest assessment by team, 
and list who failed so we can assign follow-up training.

Why it works: Pass rates by team show where the gaps are, and the list of names makes follow-up easy.

Using the SCW connector, build a leaderboard for our current Cybermon that 
I can share with the teams.

Why it works: Saying who the leaderboard is for tells the AI tool to keep it simple and shareable.

Commit and vulnerability activity

Using the SCW connector, show our commit activity for the last 90 days by 
repository. Which repositories are the most active?

Why it works: Knowing where most code is written shows where secure coding training has the biggest impact.

Using the SCW connector, show our vulnerability findings by category and 
repository, and the average time to fix them.

Why it works: Grouping findings by category and repository shows which issues come up most and where, and time to fix shows how quickly teams resolve them.

Using the SCW connector, compare our top vulnerability categories with our 
developers' training accuracy in the same categories.

Why it works: Putting findings next to training results shows where training should focus to reduce the issues you actually see.

AI usage reporting

Using the SCW connector and our AI usage data, which AI coding tools do our 
developers use, and how much of our code is AI-assisted?

Why it works: Asking about "AI usage data" points the AI tool to the right data source, and asking about tools as well as volume shows what to cover in your AI policy.

Using the SCW connector, list our top 10 AI users and their AI score.

Why it works: A top 10 keeps the list short enough to act on.

Tips for better answers

1. Start with "Using the SCW connector" so the AI tool knows where to look.
2. Name a time period, team or Quest. "In the last 90 days" works better than "recently".
3. Ask for the format you want: a table, a chart or a 3-point summary.
4. Ask follow-up questions in the same chat to dig deeper.

Related articles

• Setting up SCW's MCP Server
• What You Can Do
• SCW's MCP Server FAQ