Skip to content

uplika/uplika-plugin / skills/naver-ai-cite

naver-ai-cite

Verified

Write or rework a Naver Blog post so AI Briefing is likely to cite it, per Naver's guide and Uplika's checks. AI 브리핑 인용, 네이버 메이트.

FileKindSize
SKILL.mdskill md16.5 KB
references/official-guide.mdmarkdown8.5 KB

references/official-guide.md

View on GitHub
# Naver's guide for content that AI Briefing cites

Quoted from Naver Search's official blog (네이버 검색 공식 블로그). The Korean titles and check questions are Naver's own words, copied unchanged; the English lines are summaries. When in doubt, read the posts themselves.

| Post | Date | Link |
|---|---|---|
| AI 시대에 사용자의 선택을 받는 콘텐츠 작성 가이드 | 2026-05-26 | https://blog.naver.com/naver_search/224296857688 |
| AI 시대에 사용자의 선택을 받는 콘텐츠 작성 가이드_실전편 | 2026-06-04 | https://blog.naver.com/naver_search/224305800678 |
| 네이버 메이트 프로그램이 새롭게 시작됩니다 | 2026-05-28 | https://blog.naver.com/naver_search/224290490478 |

Each item gets pass, gap or n/a in the skill. Nothing here is a weight or a threshold; Naver publishes none.

## 콘텐츠 작성 5가지 원칙 (five principles)

Source: https://blog.naver.com/naver_search/224296857688

1. **직접 경험한 지식**: write from what you did yourself (concrete cases, how a problem was solved, failures, detailed use); on expert topics, cite statistics, research and official announcements.
2. **일관된 주제**: AI search treats a channel that keeps publishing good posts on one topic as an expert source, so fill most of the channel with the core topic you know best.
3. **거짓 없는 진정성**: say clearly, where readers see it at once, whether you were paid or given the product (원고료, 제품 협찬) or bought it yourself (내돈내산); name the author of anything you quote and link the original.
4. **읽기 쉬운 구조**: use the title, subheadings and paragraph breaks so readers find the key information fast; use images and video that fit the flow, and write their key point in text too.
5. **최신성 유지**: when information in an old post changes or a new trend appears, update the post, and keep writing new content.

## 이런 글은 사용자에게 외면 받을 수 있으니 주의해 주세요! (five things to avoid)

Source: https://blog.naver.com/naver_search/224296857688

1. **무분별한 복제와 생성**: do not copy popular posts or stitch unverified facts together, and do not publish, indiscriminately, text that an AI generated or summarized rather than a person wrote.
2. **과도한 광고와 홍보**: no ad copy or keywords unrelated to the core content, and no excessive outbound links whose only purpose is promotion.
3. **무관한 키워드 삽입**: do not force in keywords unrelated to the context just to be shown.
4. **콘텐츠와 무관한 이미지, 동영상의 삽입**: do not line up images or videos that have nothing to do with the text.
5. **부정확하고 부적절한 정보**: do not write and spread information whose facts were not checked.

## 한눈에 보는 콘텐츠 셀프 체크 가이드 (five self-checks)

Source: https://blog.naver.com/naver_search/224305800678. The questions under each title are Naver's own checklist; a pass needs a line in the draft that answers them.

1. **독자 및 목적**: write from the reader's side: who looks for this, when, in what situation, for what purpose.
   - 누가, 언제, 어떤 상황(TPO)에서, 어떤 목적으로 이 정보를 찾게 될지 독자의 관점에서 콘텐츠를 작성하셨나요?
2. **절차 및 구체적 해결 과정**: tell what you did in the order it happened, with the tips and cautions only someone who did it knows.
   - 실제 경험한 순서와 절차를 상세히 기술하였나요?
   - 창작자만이 알고 있는 팁이나 주의 사항을 포함하였나요?
3. **대안 및 비교 분석**: include the alternatives you weighed, their strengths and weaknesses, and the deciding reason, so readers can choose for their own situation.
   - 제시한 방법 외에 고려한 대안에는 어떤 것이 있고, 선택의 결정적 이유와 각 대안의 장단점을 함께 담았나요?
   - 독자의 상황에 따라 최적의 선택을 하기 위해 어떤 점들을 고려했는지 설명하였나요?
4. **실제 적용 사례 및 결과**: report what happened when you applied it, what differed from what you expected, the trial and error, and objective evidence behind the claims.
   - 선택한 방법이나 지식을 실제 적용했을 때의 결과, 예상치 못한 아쉬움이나 시행착오에 대한 후기를 담았나요?
   - 주장을 뒷받침하는 통계·연구 자료·공식 문서 등 객관적 근거를 적절히 인용하였나요?
5. **맥락에 맞는 이미지와 동영상**: place media that fits the flow (infographics, step images, short summary videos), and give its key content in text as well.
   - 콘텐츠를 이해하는 데 도움이 되도록 인포그래픽, 과정 설명용 이미지, 핵심 요약 영상 등이 적재적소에 배치되었나요?
   - 멀티미디어로 전달한 내용을 텍스트로도 함께 제공하였나요?

### Naver's good and bad examples, in short

The practical guide pairs each self-check with a good and a bad case. Summaries:

- 독자 및 목적: good names a specific reader and situation and gives the practical details they need (route, price, time taken); bad is a bare ranking list with no reader or situation.
- 절차 및 구체적 해결 과정: good follows the process in time order, with failures, fixes and measured numbers; bad is generalities, unsupported claims and unsafe advice.
- 대안 및 비교 분석: good tried several options, compares them point by point and recommends by situation; bad is generic praise anyone could write.
- 실제 적용 사례 및 결과: good states the real setup and how long it was used, with results measured or observed first-hand; bad copies the maker's spec sheet and pushes a purchase link.
- 맥락에 맞는 이미지와 동영상: good structures the key information with tables, lists and subheadings and shows photos directly tied to the text; bad drops in unrelated photos.

## What the guide's FAQ adds (paraphrased)

From the FAQ at the end of https://blog.naver.com/naver_search/224296857688:

- Content with real experience (a problem actually solved, a detailed review) is what AI Briefing cites; when a post uses images, writing their key information in text helps.
- Likely to be filtered out: repeated keyword lists, clickbait, stitched-together posts with unclear sources, and text an AI generated or summarized mechanically instead of a person writing it; heavy pop-ups, disguised ads and overly promotional posts can be left out too.
- Using AI tools is not penalized as such. The experience has to be the creator's own, and the creator should review and edit the final text.
- Ads and sponsored posts are not excluded automatically, but the financial relationship (내돈내산, 업체 지원) should be stated clearly; a body of promotional lines with no information works against the post.
- Recency is not required for every topic: real-time and trend topics favor recent posts, while for lasting information depth and structure matter more.
- Channels that have kept writing quality posts on one topic can be cited more often, and posts that already show well in search results are more likely to be cited.

## Background: 네이버 메이트 (Naver Mate)

Source: https://blog.naver.com/naver_search/224290490478 · program page https://mate.naver.com

- A content partner program for creators and communities on Naver's user-generated services (블로그, 지식iN, 카페, 프리미엄 콘텐츠), picked every month by topic, with activity support for those picked.
- Selection centers on the AI Briefing citation count (AI 브리핑 인용수), weighed with each service's own criteria. The count is how many times a creator's content was cited in AI Briefing in search results, counted cumulatively since January 2026.
- In a month a creator is picked, their content on the selected topic shows an emblem and the citation count in search results and AI Briefing.
- The program points to the two guides above for how to write content that AI Briefing cites.

## Common practice, not in Naver's text

Widely recommended for content that AI answers quote. Naver's criteria above do not include them, so always label them "not in Naver's text" when you show them.

1. **Conclusion first**: the first paragraph states the answer or the verdict, and the rest of the post supports it.
2. **Question-style subheadings**: subheadings phrased the way readers ask (예: 「숙소는 어디가 제일 편했나요?」), each answered right below.
3. **Short FAQ at the end**: a few questions readers still have, each with a short direct answer drawn from the post. (Naver's own guide ends with an FAQ, but no criterion asks for one.)
slack-gif-creatoranthropics/skillsKnowledge and utilities for creating animated GIFs optimized for Slack. Provides constraints, validation tools, and animation concepts. Use when users request animated GIFs for Slack like "make me a GIF of X doing Y for Slack."Stars: 198kUpdated Claude CodeCodexVerifiedInstall
ponytailDietrichGebert/ponytailLazy senior dev mode for any coding task: the smallest change that fully solves it, and a reply a busy human understands in one read.Stars: 196kUpdated Claude CodeCodexVerifiedInstall
ponytail-debtDietrichGebert/ponytailHarvest every shortcut: comment into one debt ledger, so deferrals get tracked instead of forgotten. One-shot report.Stars: 196kUpdated Claude CodeCodexVerifiedInstall
ponytail-debtDietrichGebert/ponytailList every `shortcut:` comment (and older `ponytail:` ones) as a debt ledger. One-shot report, changes nothing. Use for "ponytail debt", "what did ponytail defer", "list the shortcuts", /ponytail-debt.Stars: 196kUpdated Claude CodeCodexVerifiedInstall