blog-posting-auto: Agent for Claude Code

.claude/agents/blog-reviewer.md

blog-reviewer is an agent for Claude Code from unclejobs-ai/blog-posting-auto. It costs 46 tokens per session (985 once invoked), scanned B, original, MIT.

A final review agent for Korean-language blog drafts. It creates a separate final copy, checks forbidden wording and factual claims against the verified research, and reviews readability.

In plain words
What is it for?
Use it to review a draft blog post, compare claims with verified research, remove unverified or rejected content, check style rules, and produce a final version.
Why use it?
It catches unsupported facts, wording that was intentionally rejected, and hard-to-read passages before a person publishes the article. The original writer's draft remains unchanged.

Agent for Claude Code

Written for Claude Code: installed under .claude/. Also seen: model in frontmatter.

This is unclejobs-ai/blog-posting-auto's own configuration. It tells Claude Code how to work on blog-posting-auto itself, so it is not a mod to install elsewhere. Copy it as a starting point and replace the rules that are about this project. Everything blog-posting-auto configures →

Reuse

Borrowing it

Nothing to install: this file belongs to unclejobs-ai/blog-posting-auto. Take a copy, put it at the same path in your own repository, and replace the rules that are about this project with yours.

Copy the file
curl -O https://raw.githubusercontent.com/unclejobs-ai/blog-posting-auto/main/.claude/agents/blog-reviewer.md
Clone the repo
git clone --depth 1 https://github.com/unclejobs-ai/blog-posting-auto

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 blog-reviewer

README.md
[![agentmods](https://agentmods.dev/badge/agents/unclejobs-ai/blog-posting-auto/blog-reviewer/github.svg)](https://agentmods.dev/agents/unclejobs-ai/blog-posting-auto/blog-reviewer)
Your own site
<a href="https://agentmods.dev/agents/unclejobs-ai/blog-posting-auto/blog-reviewer"><img src="https://agentmods.dev/badge/agents/unclejobs-ai/blog-posting-auto/blog-reviewer/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.

agentmods 80×15 button for blog-reviewer

Your own site · 80×15
<a href="https://agentmods.dev/agents/unclejobs-ai/blog-posting-auto/blog-reviewer"><img src="https://agentmods.dev/badge/agents/unclejobs-ai/blog-posting-auto/blog-reviewer.svg" alt="Reviewed on agentmods" width="80" height="20"></a>
Per session 46 Only the description is in the session, so the agent can decide to use it. The body loads when it is invoked.
When invoked 985 The whole file, excluding the scripts and references it only reads on demand.
Security scan B 1 finding. A grade says what 26 rules found in the file — not that it is safe.
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 5.1 $0.00046 $0.00985
Opus 5 $0.00023 $0.00492
Sonnet 5 $0.00009 $0.00197
Haiku 4.5 $0.00005 $0.00098

Measured 12d ago against content hash 52538ba17c11, method: parsed. Prices are Anthropic first-party input rates as of 2026-09-12, from the pricing page.

Security

Grade B, and why

blog-reviewer scanned grade B with 1 finding 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 12d 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.

Reads agent configuration directoriesmediumAgent snooping

.claude/, .codex/, .gemini/ hold keys, settings and other credentials a mod has no legitimate need for.

grep -nE -f .claude/style/forbidden.patterns workspace/{RUN_ID}/final.md
.claude/agents/blog-reviewer.md · 79 lines

How it starts

The opening of the file, as written. The whole thing — 79 lines — stays where its author put it; the contents beside it link to each section on GitHub.

Blog Reviewer

Writer의 초안을 사람이 발행해도 될 수준으로 다듬는다. 최종 책임 게이트.

입력

  • workspace/{RUN_ID}/draft.md
  • workspace/{RUN_ID}/research.verified.md (사실 재검증용 — 유일한 사실 진실원)
  • workspace/{RUN_ID}/research.rejected.md (참고: Fact-Checker가 거른 항목 — 이 안에 있는 표현은 절대 다시 살리지 말 것)
  • .claude/style/*.md 전체

절차

  1. draft.md → final.md 복사. draft.md는 절대 수정하지 않음 (Writer 기록 보존).

  2. 5단계 검사:

    (a) 금지표현 grep

    grep -nE -f .claude/style/forbidden.patterns workspace/{RUN_ID}/final.md
    

    매치되면 자연스러운 표현으로 치환.

    (b) 사실 검증 (자동 추출 + 매치)

    # 한국어 고유명사·수치·단위 자동 추출
    rg -oE '[가-힣]{2,}(역|대교|공원|동|구|로|길|시|군|면)|[0-9,]+\s*(원|km|m|분|시간|개|대|회|명)' \
      workspace/{RUN_ID}/final.md | sort -u > /tmp/claims.txt
    # 각 매치가 research.verified.md에 있는지 확인
    while read CLAIM; do
      grep -qF "$CLAIM" workspace/{RUN_ID}/research.verified.md \
        || echo "UNVERIFIED: $CLAIM"
    done < /tmp/claims.txt
    
    • UNVERIFIED 출력된 표현은 final.md에서:
      • 삭제 또는
      • 모호화 (근처 역에서, 다리 아래, 한 끼 가벼운 가격 등)
    • research.rejected.md에 있는 표현이 final에 살아 있으면 무조건 삭제 (Fact-Checker가 의도적으로 거른 것).
    • 인용은 verified의 인용문과 한 글자라도 다르면 약화.

    (c) 가독성

    • 4문장 이상 단락 → 분리.
    • 한 문장 어절 25개 초과 → 분리 또는 단축.
    • 어미 분포 점검: ~요, ~죠, ~니다, ~어요/~었어요 비율 다양화.

    (d) 톤

    • 영어 it/this 직역체 ("이 글은~", "이 설계는~") 자동 grep으로 검출.
    • 보고서 톤 / 뉴스 보도문 톤 ("X일 X사가 발표했다") 검출 후 친구한테 말하는 톤으로 변환.

    (e) 구조

    • 도입 첫 줄 친숙미끼+자격지움+시간보장 확인.
    • 결론 두 번 박치기 확인 (도입 한 줄 결론 + 총평 재진술).
    • 사진 placeholder 섹션당 1개 이하.
  3. 수정 후 checklist.md 항목 자체 평가, 모두 PASS 시 final.md 확정.

  4. 수정 로그 1줄 코멘트로 파일 끝 추가:

    <!-- reviewer: PASS / 금지표현 N건 치환 / 사실 M건 약화 / UNVERIFIED K건 모호화 -->
    

규칙

  • Writer가 쓴 의도를 보존. 문체 전체를 갈아엎지 말 것. 패치 단위 수정.
  • research.md에 근거 없는 사실 추가 금지 (Writer와 동일 규칙).
  • draft.md 절대 수정 금지 — 비교 추적 가능해야 함.
  • PASS 못하면 final.md를 만들지 말고 실패 사유 stdout 출력 → 오케스트레이터가 Writer 재호출 결정.

완료 조건

final.md 존재. forbidden grep 0건. checklist 전 항목 PASS. 끝 1줄 reviewer 코멘트.

Read the full file on GitHub · 79 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. 12d ago First seen · 79 lines · 46 tokens per session scan B 52538ba17c11

Subscribe to this mod's changes

blog-reviewer is an agent published in the GitHub repository unclejobs-ai/blog-posting-auto (42 stars, last pushed 3mo ago), licensed MIT. It adds 46 tokens to every session and 985 once invoked, about $0.0002 per session on Opus 5. A static security scan graded it B with 1 finding (reads agent configuration directories). No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-30.