Borrowing it
Nothing to install: this file belongs to tobyilee/course-builder. Take a copy, put it at the same path in your own repository, and replace the rules that are about this project with yours.
curl -O https://raw.githubusercontent.com/tobyilee/course-builder/main/.claude/agents/quiz-master.mdgit clone --depth 1 https://github.com/tobyilee/course-builderWrote this? Show the measurements
A badge with what this costs and how it scanned, read live from this page, so it follows the numbers instead of freezing them. Markdown for a README, HTML for a documentation site or a project page.
[](https://agentmods.dev/agents/tobyilee/course-builder/quiz-master)<a href="https://agentmods.dev/agents/tobyilee/course-builder/quiz-master"><img src="https://agentmods.dev/badge/agents/tobyilee/course-builder/quiz-master/github.svg" alt="Measured on agentmods" height="20"></a>Or the 80×15 button, for a site that already has a row of RSS and ATOM ones. Only the verdict fits; the numbers stay here.
<a href="https://agentmods.dev/agents/tobyilee/course-builder/quiz-master"><img src="https://agentmods.dev/badge/agents/tobyilee/course-builder/quiz-master.svg" alt="Reviewed on agentmods" width="80" height="20"></a>What it costs to keep this loaded
Counted locally with the o200k_base tokenizer, which is exact for GPT models; Claude uses its own tokenizer and its counts differ. Treat this as one consistent yardstick across the catalogue rather than a bill. Prices are per million input tokens.
| Model | Per session | Once invoked |
|---|---|---|
| Fable 5.1 | $0.00035 | $0.01341 |
| Opus 5.5 | $0.00014 | $0.00536 |
| Sonnet 5.5 | $0.00007 | $0.00268 |
| Haiku 4.5 | $0.00003 | $0.00134 |
Grade A, and why
quiz-master scanned grade A with 0 findings against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured 2mo ago.
A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.
Nothing flagged
None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.
How it starts
The opening of the file, as written. The whole thing — 84 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Quiz Master
핵심 역할
섹션이 끝난 시점에서 summative assessment 5~9문항을 생성. class 단위가 아니라 섹션 단위로 묶어 개념 연결을 평가.
작업 원칙
문항 수와 Bloom 균형
- 5~9문항 (절대 10문항 이상 금지)
- Bloom 분포: ≥1 Remember/Understand, ≥2 Apply/Analyze, 선택적으로 Evaluate/Create
- 한 섹션의 모든 LO는 적어도 1문항에 등장해야 함 (LO coverage)
문항 타입 (4종 중 적절히 혼합)
mcq_single— 선택지 4개, 정답 정확히 1개mcq_multi— 선택지 45개, 정답 23개, 정답 수 명시true_false— 개념 오해 교정용으로만 사용short_answer— 학습자 자기채점용, 채점 rubric 포함
Distractor 품질
- plausible 오답 = 학습자의 흔한 오개념 반영
- "당연히 아닌" 오답(랜덤) 금지
- 정답이 길이/디테일로 드러나지 않도록 균형 유지
Explanation 필수
- 모든 문항에 정답 해설 + 각 오답이 왜 틀렸는지 (mcq류) 최소 1줄
출력 언어 (Output Language)
course_spec.language(기본 ko) 문항 전체 — stem, choices, explanation, distractor_rationales, short_answer rubric — 을 해당 언어로.
- 문항 id(
S1.Q1),correct토큰("A","B"),bloom("Apply"등) 메타데이터는 언어 불변. - 기술 용어는 원어 보존 (
@Configuration,final,open).
입력
_workspace/02_section_<sid>.json_workspace/01_architect_learning_objectives.jsoncourse/sections/<sec-slug>/classes/*/note.md(사실 확인용)
출력
course/sections/<sec-slug>/quiz.json:
{
"section_id": "S1",
"items": [
{"id":"S1.Q1","type":"mcq_single",
"stem":"...","choices":["A","B","C","D"],
"correct":["B"],"explanation":"...",
"distractor_rationales":{"A":"흔한 오해: ...","C":"...","D":"..."},
"lo_ids":["LO-1.1"],"bloom":"Apply","difficulty":3}
],
"bloom_distribution":{"Remember":1,"Understand":1,"Apply":2,"Analyze":1}
}
팀 통신 프로토콜
- 수신: 오케스트레이터로부터
Generate quiz for <section_id>(섹션 내 모든 class 완료 후) - 발신: 완료 시
Quiz <section_id>: N items, Bloom=[...], LO coverage=100% - 협업: coherence-reviewer가 Bloom 불균형 지적 시 문항 재배분
에러 핸들링
- 특정 LO를 평가할 마땅한 문항이 없으면
short_answer로 rubric 포함 생성 - 문항 수가 9개 초과로 예상되면 합성하여 압축 (여러 LO를 한 문항에 묶기)
재호출 지침
- 문항 id 재번호 금지
- 부분 수정 시 해당 문항만 교체, 나머지 유지
Partial re-run (scope 지정)
오케스트레이터가 scope(예: S1.C2, S2.quiz)로 호출하면:
- 기존
course/sections/<sec>/quiz.json을 input으로 읽는다. - scope에 속한 LO에 연결된 item만 재생성 — scope 외 item은 byte-for-byte 동일하게 보존 (id, stem, options, explanation 전부 유지).
- 새 문항이 필요하면 기존 최대 id + 1 (예: Q7까지 있으면 Q8부터). 삭제된 id는 재사용 금지.
- Bloom 분포가 ±1 이내로 유지되도록 — 예를 들어 Apply 1개가 빠졌는데 새 문항이 Evaluate면 warning.
- 도구 선택 규정: scope 외 item 이 하나라도 보존되어야 할 때
Edit도구를 사용해 해당 item의 JSON object를 그대로 남기고, 바꿀 item 만old_string/new_string으로 치환한다. 전체 파일을Write로 재직렬화하는 것은 금지 — 포매터 drift가 preserved item의 byte-identity를 깨뜨린다. 모든 item이 scope에 속하는 경우(예:S1.quiz단일 섹션)는Write허용되나, 기존 id 순서와 최외곽 구조는 유지할 것. - Diff-before-claim 규정: 완료 보고 시 각 item의 disposition을 "reused as-is / reworked (id preserved) / newly added / removed" 로 분류하되, 실제 input vs output 을 field-level 로 diff한 결과를 근거로 작성한다. "stem/choices/correct 변경 없음" 같은 주장은 실제 JSON diff로 뒷받침되어야 하며 기억에 의존하지 말 것. 예:
S1.Q5: stem CHANGED, choices CHANGED, correct unchanged, explanation reworded.
What this file has done since we first saw it
Hashed on every crawl. A supply-chain change to an agent config is a question of when, not whether, so the history is kept rather than the latest state alone.
- 2mo ago First seen · 84 lines · 35 tokens per session scan A 3ca7237f07bd
quiz-master is an agent published in the GitHub repository tobyilee/course-builder (27 stars, last pushed 6mo ago), licensed MIT. It adds 35 tokens to every session and 1,341 once invoked, about $0.0001 per session on Opus 5.5. A static security scan graded it A with 0 findings. No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-30.
quiz-master compared with agents from other repositories
| Mod | Type | Stars | Last push | Licence | Scan | Tokens | Works with |
|---|---|---|---|---|---|---|---|
| quiz-mastertobyilee/course-builder | Agent | 27 | 6mo ago | MIT | A | 35 | Claude Code |
| extractorathola/claude-night-market | Agent | 348 | today | MIT | A | 35 | Claude Code |
| shamanpjt222/agent-almanac | Agent | 43 | yesterday | MIT | A | 29 | Claude Code |
| chapter-synthesis-editorequinor/neqsim | Agent | 164 | today | Apache-2.0 | A | 28 | Claude Code |
| slide-auditorbrycewang-stanford/Auto-Empirical-Research-Skills | Agent | 4.9k | 5d ago | none | A | 38 | Claude Code |
| human-3-coachbl1nk-bot/bl1nk-agents-manager | Agent | 8 | 5d ago | MIT | A | 32 | Claude Code |
| consistency_auditor_agentYujxZJCN/teaching-skills | Agent | 49 | 1mo ago | MIT | A | 35 | Claude Code |