Create strict JSON single-answer multiple-choice quizzes.
80%
Does it follow best practices?
Run evals on this skill
Adds up to 20 points to the overall score
Passed
No findings from the security scan
Fix and improve this skill with Tessl
tessl review fix ./docker/sandbox/skills/quiz/SKILL.mdCreate one /workspace/<slug>.json file. JSON is the canonical quiz; do not
write a separate Markdown summary.
Use exactly this closed version-one shape:
{
"schema_version": 1,
"title": "Quiz title",
"questions": [
{
"question_text": "One focused question",
"options": ["Option A", "Option B", "Option C", "Option D"],
"correct_option_index": 1,
"explanation_text": "Why Option B is correct."
}
]
}correct_option_index is an integer from 0 through 3.explanation_text is required and should concisely teach why the selected
answer is correct.\(...\) for
inline math and \[...\] for display math. Escape each backslash as \\ in
JSON. Keep delimiters and braces balanced and do not nest math delimiters.schema_version, use Markdown or raw HTML, include
links or media, duplicate question text, or use null.Call:
verify_artifact(path="/workspace/<slug>.json", format="quiz")Fix all blocking findings together, rewrite the JSON once, and reverify the exact path. If a blocker remains, stop and explain it instead of looping.
After verification returns status="verified", call:
save_artifact(path="/workspace/<slug>.json", title="...")Do not pass markdown_representation; the backend derives trusted Markdown
from the exact receipt-bound JSON.
For revisions, edit the restored JSON from primary_path, write the complete
quiz to expected_output_path, verify that path as quiz, and save with the
returned artifact_id and expected_generation. Do not edit the derived
markdown_path or reconstruct the quiz with vision.
MODSetter/SurfSense
e3de9e0
If you maintain this skill, you can claim it as your own. Once claimed, you can manage eval scenarios, bundle related skills, attach documentation or rules, and ensure cross-agent compatibility.