agentleFS
williamcorrea23williamcorrea23/sap-router-skill/.agents/skills/cap-trivia/SKILL.md

cap-trivia

Skill1 starsChanged 4 months agoMITView on GitHub

Run an interactive CAP trivia quiz. Picks a topic, searches the CAP documentation via MCP, and asks multiple-choice questions using the question tool. Optionally accepts a topic and difficulty. Use when a developer wants to test or deepen their CAP knowledge.

1 other project uses a version of this.

Adopted by

Most recently taken up by

The file

Show its rules and text

We split this skill into the rules it gives an agent

14 rules. About 0 are common, 9 are distinctive, and 5 are very specific to this project.

  • ≈0

    Common · estimate

    Many files like this have a rule of this kind.

  • ≈9

    Distinctive · estimate

    This author's own choices. Worth a look.

  • 5

    Very specific

    Tied to their project's names, files or tools. Probably skip.

  • Common
  • Distinctive
  • Very specific
  1. Very specific:

    If neither is given, pick a topic at random from the list above and choose intermediate difficulty.

    line 25
  2. Distinctive:

    If the user provided a topic and/or difficulty, use those.

    line 32
  3. Very specific:

    Use the cds-mcp_search_docs tool to find accurate, specific content for the chosen topic.

    line 37
  4. Distinctive:

    Write exactly 3 multiple-choice questions grounded in what the documentation says.

    line 54
  5. Distinctive:

    Avoid questions with answers like "all of the above" or "none of the above"

    line 70
  6. Very specific:

    Call the question tool once with all questions as an array.

    line 74
  7. Distinctive:

    After receiving the answers, output a compact result — one line per question, then the final score:

    line 85
  8. Distinctive:

    Only add extra explanation for wrong answers.

    line 96
  9. Distinctive:

    If the user scored less than 60%, suggest they read the relevant docs section and offer to run another round on the same topic at a lower …

    line 99
  10. Distinctive:

    If the user scored 100%, offer a harder round on the same topic or a new topic.

    line 102
  11. Distinctive:

    Always use the MCP doc search to source questions — do not rely on training knowledge alone.

    line 118
  12. Very specific:

    Prefer questions about Node.js behavior unless the user specifically asks for Java.

    line 120
  13. Very specific:

    Always ask exactly 3 questions.

    line 122
  14. Distinctive:

    If the user provides a topic you can't find good doc material for, say so and suggest a nearby topic you can cover well.

    line 124

≈ “Common” is an estimate: at least 1 in 5 of the 350,495 skill files we read have a rule of the same kind, not the same words. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.

Keep what fits and drop the rest. Your version, ready for your agents.
Files
SKILL.md

What I do

Generate a short interactive CAP trivia quiz. Pick (or receive) a topic, search the CAP docs for accurate source material, write 3 multiple-choice questions, and present them to the user using the question tool so they appear as interactive choices — not just text.

Inputs (optional)

The user may provide:

  • topic — e.g. "CDS modeling", "authorization", "databases", "MTX", "OData", "event queues", "handlers", "testing"
  • difficulty — beginner, intermediate, or advanced

If neither is given, pick a topic at random from the list above and choose intermediate difficulty.

Steps

1. Resolve topic and difficulty

If the user provided a topic and/or difficulty, use those. Otherwise, pick randomly. Do not output anything at this step — go straight to searching.

2. Search the CAP docs

Use the cds-mcp_search_docs tool to find accurate, specific content for the chosen topic. Run 2–3 searches to gather enough material for varied questions. Good search queries:

TopicExample search queries
CDS modeling"entity aspects associations managed", "localized temporal"
Authorization"restrict requires roles", "instance-based authorization"
Databases"sqlite hana postgres plugin cap-js", "schema evolution deploy"
OData"odata draft enabled fiori", "odata actions functions"
Event queues"cds queued transactional event queue inbox"
Handlers"srv.on before after handler phases", "req.reject error"
Testing"cds.test jest supertest cap"
MTX"cds-mtxs multitenancy sidecar extensibility"

3. Compose the questions

Write exactly 3 multiple-choice questions grounded in what the documentation says. Each question must have 4 answer options (A–D).

Guidelines by difficulty:

DifficultyFocus
beginnerBasic concepts, CLI commands, what things are called
intermediateHow things work, when to use which approach, config
advancedEdge cases, subtle behavior, annotations with specific rules, internal mechanics

Rules for good questions:

  • Base every question on something found in the docs — no guessing
  • Make exactly one option clearly correct
  • Make the distractors plausible but wrong (not obviously silly)
  • Vary the format: "which of these", "what happens when", "which annotation"
  • Avoid questions with answers like "all of the above" or "none of the above"

4. Present questions using the question tool

Call the question tool once with all questions as an array. Each question should have:

  • header: very short label, e.g. "Q1", "Q2"
  • question: the full question text
  • options: four choices, each with a concise label (the answer text) and a description that adds a brief hint or context

The user selects their answers and submits all at once.

5. Score and explain

After receiving the answers, output a compact result — one line per question, then the final score:

Q1 ✅  @odata.draft.enabled — enables OData draft for Fiori edit flows.
Q2 ❌  Correct: srv.before — runs before the generic handler, not after.
Q3 ✅  @restrict — declarative authorization via roles and conditions.

Score: 2 / 3

Only add extra explanation for wrong answers. Keep correct answers to one short sentence.

If the user scored less than 60%, suggest they read the relevant docs section and offer to run another round on the same topic at a lower difficulty.

If the user scored 100%, offer a harder round on the same topic or a new topic.

Example question shape

Q3: Which annotation enables OData draft support for an entity?

A) @readonly
B) @odata.draft.enabled    ← correct
C) @cds.persistence.skip
D) @Capabilities.Updatable

Notes

  • Always use the MCP doc search to source questions — do not rely on training knowledge alone. CAP APIs and defaults change across major versions.
  • Prefer questions about Node.js behavior unless the user specifically asks for Java.
  • Always ask exactly 3 questions. No upfront announcements, no preamble — just search, then present the questions via the question tool.
  • If the user provides a topic you can't find good doc material for, say so and suggest a nearby topic you can cover well.

Download now (ZIP)

Does it work?

Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.

No reviews yet

People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.

  • It worked
  • Partly
  • It didn't
What did you use it for? e.g. Reviewing pull requests in a Rails app

Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.

Your trail

Files you open will collect here, so you can find them again.

Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.

Now do this for your own.

Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.