coherence-reviewer

coherence-reviewer is an agent for Claude Code from tobyilee/course-builder. It costs 45 tokens per session (1,612 once invoked), scanned A, original, MIT.

A quality checker for a course’s slides, notes, spoken script, and quizzes. It checks whether these materials cover the same learning goals and tell a consistent story.

In plain words
What is it for?
Use it to review course materials section by section and return a pass or revise decision. When revisions are needed, it identifies which agent should make them and what to change.
Why use it?
It finds missing learning goals, mismatched slide references, conflicting quiz answers, uneven difficulty, tone changes, and script content that is hard to say aloud before the course is released.

Agent for Claude Code

Install

Getting it into your agent

One page per mod, every tool's command on it. A separate URL per tool would split the same page into five that compete with each other.

agentmods
npx agentmods add agents/tobyilee/course-builder/coherence-reviewer
Clone the repo
git clone --depth 1 https://github.com/tobyilee/course-builder

Made for: Claude Code.

Wrote 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.

agentmods badge for coherence-reviewer

README.md
[![agentmods](https://agentmods.dev/badge/agents/tobyilee/course-builder/coherence-reviewer.svg)](https://agentmods.dev/agents/tobyilee/course-builder/coherence-reviewer)
Your own site
<a href="https://agentmods.dev/agents/tobyilee/course-builder/coherence-reviewer"><img src="https://agentmods.dev/badge/agents/tobyilee/course-builder/coherence-reviewer.svg" alt="Measured on agentmods" height="20"></a>
Per session 45 Only the description is in the session, so the agent can decide to use it. The body loads when it is invoked.
When invoked 1,612 The whole file, excluding the scripts and references it only reads on demand.
Security scan A 0 findings. Scan, not verified.
Origin original No closer match found in the catalogue.
Token cost

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.

ModelPer sessionOnce invoked
Fable 5 $0.00045 $0.01612
Opus 5 $0.00023 $0.00806
Sonnet 5 $0.00009 $0.00322
Haiku 4.5 $0.00005 $0.00161

Measured 4d ago against content hash d867fb8485a4, method: parsed. Prices are Anthropic first-party input rates as of 2026-08-30, from the pricing page.

Security

Grade A, and why

coherence-reviewer 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 4d 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.

.claude/agents/coherence-reviewer.md · 125 lines

How it starts

The opening of the file, as written. The whole thing — 125 lines — stays where its author put it; the contents beside it link to each section on GitHub.

Coherence Reviewer

핵심 역할

하네스의 QA 게이트. 9개 자산이 서로 정합한지 경계면 비교로 검증. 단순히 존재 확인이 아니라 동일 LO에 대해 slide/note/transcript/quiz가 같은 스토리를 말하는가를 본다.

작업 원칙

검사 항목 (fail-fast 순서)

  1. LO Coverage

    • 모든 LO가 ≥1 slide, ≥1 note 문단, ≥1 quiz item에 등장하는가
    • 한 번도 평가되지 않는 LO는 FAIL
  2. Bloom Balance

    • 섹션당 Bloom 레벨 ≥3개 분포
    • Course 전체 ≥4개 레벨 커버
  3. Slide ↔ Script 정합

    • slide N개 vs script의 [slide N] cue 개수 일치
    • slide.source.md에 없는 슬라이드를 script가 지칭하면 FAIL
  4. Note ↔ Slide 정합

    • note의 [slide N] cross-ref가 실제 슬라이드 번호 범위 내
    • note가 slide에 없는 개념을 새로 도입하면 경고
  5. Tone 일관성

    • 한 class 내 slide/note/transcript의 톤 파라미터(friendly/formal/socratic) 준수
    • 문체 급변(반말 ↔ 존댓말 혼재 등) 검출
  6. Quiz 사실 검증

    • quiz 정답이 note/slide 내용과 모순되지 않는가
    • distractor가 정답과 구분 가능한가
  7. Speakability

    • transcript에 raw code literal / 길이 ≥15의 숫자 / 생 URL 있으면 FAIL

판정 결과

  • pass / revise 2단계
  • revise 시 수정 대상 에이전트와 구체 지침을 메시지로 전달 (SendMessage)

입력

  • course/ 하위 모든 자산
  • _workspace/ 하위 모든 spec

출력

원칙: atomic dual-write. 매 검사 라운드마다 JSON과 MD를 동일 소스 데이터에서 한 번에 생성한다. 중간 저장 후 한 쪽만 overwrite하지 말 것.

  • _workspace/99_coherence_report.json — machine contract (build gate가 이 파일 .overall만 읽음)
  • _workspace/99_coherence_report.md — 사람이 읽는 요약

Verdict 동기화 (필수)

JSON overall 값과 MD ## VERDICT: H2 헤더는 항상 동일한 verdict를 표시해야 한다. 대소문자는 case-insensitive 매칭 (build-bundle.sh 도 lowercase로 정규화 후 비교). 가독성을 위해 MD는 대문자 권장이지만 lowercase 도 contract 준수로 인정:

JSON overall MD 헤더 (권장) MD 헤더 (허용)
"pass" ## VERDICT: PASS ## VERDICT: pass
"revise" ## VERDICT: REVISE ## VERDICT: revise

중요한 건 JSON .overall 과 MD 헤더가 같은 verdict 의미를 갖는 것 — 한쪽이 pass 인데 다른쪽이 revise 면 contract 위반.

Revise 루프 종료 처리

수정 라운드 완료 후 최종 verdict가 PASS로 바뀌면:

  1. 반드시 두 파일 모두 최종 verdict로 overwrite. 이전 라운드의 REVISE MD가 남아있으면 gate는 pass하지만 사람이 읽는 리포트는 거짓을 말함.
  2. 이전 라운드를 보존하고 싶으면 99_coherence_report.round_<N>.{json,md} 로 분리해 저장.
  3. JSON 안 prior_revisions[] 에 해결된 이슈 기록 + MD 본문에도 동일 섹션 추가.

Read the full file on GitHub · 125 lines

Changes

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.

  1. 4d ago First seen · 125 lines · 45 tokens per session scan A d867fb8485a4

Subscribe to this mod's changes

coherence-reviewer is an agent published in the GitHub repository tobyilee/course-builder (22 stars, last pushed 4mo ago), licensed MIT. It adds 45 tokens to every session and 1,612 once invoked, about $0.0002 per session on Opus 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.

Related

Other agents, from other repositories

artifact-coverage-reviewer

Independent post-finalization coverage reviewer. Walks every ## Verification Notes and ## Precedents & Lessons entry in a finalized artifact and verifies each lands somewhere actionable — either reflected in a phase's ### Success Criteria: bullet or visibly addressed by the slice's emitted code. Emits one…

juicesharp/rpiv-mono · 102 tokens

demo-producer

Universal demo video producer that creates polished marketing videos for any content - skills, agents, plugins, tutorials, CLI tools, or code walkthroughs. Uses VHS terminal recording and Remotion composition.

yonatangross/orchestkit · 42 tokens

redaccion

Eres un experto en redacción académica en LaTeX para Trabajos de Fin de Grado (TFG) y Máster (TFM) de la Escuela Politécnica Superior (EPS) de la Universidad de Alicante (UA).

jmrplens/TFG-TFM_EPS · 0 tokens

prompting-tutorials

This page documents the best-performing LLM prompts for creating SolidWorks parts via the MCP server. Each recipe shows the exact sequence of tool calls and the prose prompt that reliably produces them from a general-purpose LLM (Claude, GPT-4o, etc.).

andrewbartels1/SolidworksMCP-python · 0 tokens

(@_@) engineer

Technical feasibility assessment, architecture review, and implementation complexity analysis. Use when evaluating technical specs, reviewing PRDs for engineering feasibility, estimating implementation effort, or getting feedback on system design decisions.

carlvellotti/free-ai-courses · 43 tokens

evidence-cohort-teacher

Produces one bounded, review-only SKILL.md body proposal from an already-selected local evidence cohort.

selftune-dev/selftune · 26 tokens