report
Skill0 starsChanged 13 days agoMITView on GitHub
Write a Korean analysis report as a Word (.docx) file in the CUBRID house style, with charts, diagrams, and color-coded result tables. Use for analysis, comparison, verification, or status reports on Hibernate, JDBC, or CUBRID work, for the team or the team lead. For Markdown notes, use note or worklog. Triggers on phrases like '보고서 만들어', 'Word 문서로 정리', '보고용 문서 작성'.
93 other projects use a version of this.
Adopted by
Most recently taken up by
- BermanRoman/promt-seo-geotheir own version, first seen Nov 8, 2026
- eliasjnior/claude-skillstheir own version, first seen Nov 8, 2026
- ai-search-guru/getcito-worlds-first-open-source-aio-aeo-or-geo-tooltheir own version, first seen Nov 8, 2026
- leinardi/adversarial-review-looptheir own version, first seen Nov 8, 2026
- KadirHarmanc/nazar-skilltheir own version, first seen Nov 8, 2026
- zedarvates/frugal-skillstheir own version, first seen Nov 8, 2026
Used with
Projects with it also use opportunities, export and 10 more.
Other files people use alongside this one.
Files found in the same public projects as a version of this one, more often than they turn up anywhere else. Each project counts once, large collections count for less, and this file's own repository is left out.
ansvisorSkillopportunities
Finds content gaps, prompts worth tracking, and the off-site sources AI cites instead of the brand.
Also in 3 projects that use this
soilmassSkillexport
Animation export configurations including format selection, resolution, framerate, quality, and optimization settings. Use when rendering animations for web, mobile, video, or social media. Provides tokens for all export parameters across Lottie, GIF, video, and sprite sheet formats.
Also in 4 projects that use this
GoogleSkillquality
Evaluates whether a GitHub issue is spam, empty, needs more information, or is OK to proceed.
Also in 3 projects that use this
genomewalkerSkillresume
Resume a thread by loading its ~800-token context capsule
Also in 7 projects that use this
tokyubevoxelverseSkillcheckup
Give a repository a full health exam — doc rot, dependency staleness, CI decay, git hygiene, dead links, docs contradicting code — and grade it. Use when the user wants to know what quietly rotted, or a prioritized maintenance list.
Also in 3 projects that use this
rajnavakotiSkillscan
Run the AI dashboard scanner and generate the HTML report.
Also in 5 projects that use this
Other versions93 projects use a version of this.
Word for word · 0
None found.
Edited · 0
None found.
Wrote their own version · 93
- microsoft/data-formulator★ 17ksame name
- Zafer-Liu/Data-Analysis-Agent★ 2.6ksame name
- H-mmer/pentest-agents★ 815same name
- ai-search-guru/getcito-worlds-first-open-source-aio-aeo-or-geo-tool★ 432same name
- cwinvestments/memstack★ 425same name
- automagik-dev/genie★ 351same name
- namastexlabs/automagik-genie★ 351same name
- ansvisor/ansvisor★ 233same name
- andrew-yangy/gru-ai★ 153same name
- zhaji2333/CkSKILLS★ 145same name
83 more not listed
How matches are found. Word for word: the exact same text. Edited: shares passages with this file (“99% same” means 99% of its five-word phrases also appear here; 10% or more counts). Their own version: similar meaning, or, weakest, the same name. We can’t tell which project had it first. Projects are counted among the 200 most-starred candidates, and mentions are links to the repository we found, up to 50 per site, so both counts are a floor.
In this repository22 other files it gives its agents.
- CLAUDE.mdCLAUDE.md
- blogSkill
- ci-checkSkill
- cmt-manualSkill
- commitSkill
- cubrid-manualSkill
- hhh-dialect-verify-reportSkill
- hhh-dialect-verifySkill
- java-formatSkill
- jdbc-ctp-verify-reportSkill
- jira-draftSkill
- jira-fetchSkill
- noteSkill
- pr-draftSkill
- pr-reviewSkill
- review-triageSkill
- self-reviewSkill
- skill-createSkill
- submodule-bumpSkill
- tutorSkill
- weekly-reportSkill
- worklogSkill
The file
We split this skill into the rules it gives an agent
6 rules. 0 are common, 2 are distinctive, and 4 are very specific to this project.
0
Common
Most of the 68 edited versions we read keep these.
2
Distinctive
This author's own choices. Worth a look.
4
Very specific
Tied to their project's names, files or tools. Probably skip.
- Common
- Distinctive
- Very specific
- Very specific: line 45
No em-dash: never use the — character in the document.
- Very specific: line 49
Write <topic>.json. See assets/example.json for a complete example.
- Distinctive: line 70
Follow the type skeleton from Step 1 and the 작성 원칙 above.
- Very specific: line 92
Font: put font-family="'맑은 고딕','Malgun Gothic','Apple SD Gothic Neo',sans-serif" on every <text>.
- Distinctive: line 96
Spacing: leave ≥ 24px between boxes.
- Very specific: line 97
Scope: use SVG for schematic diagrams (boxes + arrows + short labels).
“Common” is measured: kept by at least 6 in 10 of the 68 edited versions of this file we could read. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.
Files
Analysis report → Word (.docx)
Generate a CUBRID-house-style Korean Word report by writing a JSON spec and running the bundled generator (assets/build_report.js), which assembles the .docx with docx-js (the engine behind Anthropic's docx skill) and embeds matplotlib charts. The exact house design is reproduced:
- Centered cover: title (bold 23pt navy #1F3864), subtitle (12pt gray), meta line (9pt gray), bold conclusion abstract (10pt)
- Auto table of contents generated from the
h1sections - Numbered headings: Heading 1 = navy 15pt bold, Heading 2 = blue 12pt
- Bordered tables with row-level color coding: sky-blue (#D5E8F0) header row, green (#E2EFDA) pass rows, red (#F8D7DA) fail rows, gray (#F2F2F2) neutral-emphasis
- Running header set to the report title; footer page numbers
Step 0: Dependencies
assets/build.sh installs what is missing on its own (global docx, the matplotlib venv), so there is no separate setup step. The document is assembled with docx-js; chart/diagram blocks are rendered by assets/figures.py (matplotlib) and embedded as images. Extra tools, needed only when those blocks are used: a mermaid block needs curl + network to reach Kroki (or set KROKI_URL to a self-hosted instance; if Kroki is unreachable the block degrades to a code block and the build continues); svg blocks and the Step 4 visual verification need LibreOffice (soffice) and pdftoppm (poppler).
Step 1: Identify the report type and gather inputs
| Type | Body skeleton |
| 에러 분석 | 증상 → 재현 → 원인 → 영향 → 해결 |
| 코드 분석 | 대상 → 구조/흐름 → 발견사항 → 개선안 |
| 이슈 분석 | 배경 → 현황 → 원인 → 해결방향 → 계획 |
| 비교 분석 | 기준/대상 → 항목별 비교표 → 권고 |
| 조사/현황 | 개요 → 방법 → 결과(표) → 해석 → 다음 단계 |
Collect (ask only for what is missing): title, subtitle (scope/version), meta (author/team · date: author defaults to CUBRID Dev1; do not ask for it unless the user names a different author), the headline conclusion, body content, and table data.
작성 원칙: write less, show more
Optimize for a reader who skims. Keep prose minimal and let structure carry the detail:
- Lead with the conclusion: the cover
conclusionstates the answer first;1. 개요is one short paragraph. - Core points only: short sentences, one idea per bullet; cut background the reader can infer.
- Prefer tables / charts / figures over paragraphs: turn comparisons, metrics, and status into a color-coded
table; turn trends or distributions into a chart embedded via animage; use anotebox for the single most important caveat. Reservepparagraphs for the few sentences that truly need prose. - One fact per bullet, and no long inline lists: do not pack
현재 → 목표into one sentence (make it two table columns), and never chain more than four names (classes, files, options) with commas. Split them into atablewhen the reader needs each name, or compress to 기준 + 개수 when they only need the scope (e.g. "Wrapper를 구현하는 7개 클래스"). - Emphasize only the few key terms with
**…**so the eye lands on them. - No em-dash: never use the
—character in the document; use commas, colons, parentheses, or periods instead.
Step 2: Write the JSON spec
Write <topic>.json. See assets/example.json for a complete example. Schema:
title,subtitle,meta(작성자/팀 · 작성일 YYYY-MM-DD: author/team + date only; author defaults toCUBRID Dev1unless the user specifies another; no org name, scope, or version on this line: those go in the subtitle or 부록),conclusion(bold cover abstract),header(optional; defaults to title),auto_number(optional:true→ h1 sections auto-numbered1.,2., … and TOC stays in sync; then give h1 bare titles without numbers).blocks: an ordered list of:{"t":"h1","text":"1. 개요"}: numbered section (TOC auto-built from these; number them1.,2., …){"t":"h2","text":"6.1 …"}: subsection{"t":"p","text":"…","bold":false}: paragraph (1.15 line spacing). Use**핵심어**anywhere to emphasize a phrase in navy bold.{"t":"ul","items":["…","…"]}: bullet list ('•'){"t":"table","header":[…],"aligns":["left","center", …],"rows":[{"cells":[…],"status":"good"}, …]}: default alignment: col0 left, other columns centered{"t":"note","text":"강조할 핵심/주의","kind":"info|warn|bad"}: shaded callout box (참고/주의/경고){"t":"image","path":"figure.png","caption":"그림 1","width_in":6.0}: embed any existing image, centered + caption{"t":"chart","kind":"bar","title":"…","subtitle":"…","note":"하단 주석","signed":false,"bars":[{"label":"…","value":N,"color":"base|blue|good|lightgreen|warn|bad","badge":"강조\n둘째 줄","note":"바 위 메모"}]}: vertical bars: bold-navy value labels, optional pillbadge/ colorednoteabove a bar, subtle baseline (house style, no axes/grid){"t":"chart","kind":"hbar","title":"…","subtitle":"…","note":"…","bars":[{"label":"…","value":N,"color":"…","tag":"유지","tag_color":"good"}]}: horizontal ranked bars: label left (+ optional coloredtag), proportional bar, bold-navy value at the end. Best for ranking magnitudes (signeddefaults true)- For trends or share:
{"t":"chart","kind":"line|pie","labels":[…],"series":[{"name":"…","data":[…]}]} {"t":"mermaid","code":"flowchart LR\n A[…] --> B{…}","caption":"그림 1","width_in":6.4}: Mermaid diagram: PREFER THIS for every structured 도식 (flow/workflow, sequence, state, ER, class, architecture). Nodes auto-size to their text, so labels never clip. Rendered to PNG via Kroki (server-side headless browser) and embedded, so Word/LibreOffice show text + fills faithfully. Follow the Mermaid 도식 작성 규칙 below. (Needscurl+ network, or a self-hostedKROKI_URL; if Kroki is unreachable the block degrades to a code block, so the build never fails on this.){"t":"svg","code":"<svg …>…</svg>","caption":"그림 2","width_in":6.4}: hand-authored SVG, embedded as a native vector image with a PNG fallback. Use only for bespoke visuals a standard Mermaid diagram can't express (custom geometry, annotated layouts, non-graph illustrations). Follow the SVG 도식 작성 규칙 below. (Needs LibreOffice for the fallback.){"t":"diagram","direction":"LR|TB","nodes":[…],"edges":[…],"caption":"그림 2"}: (legacy) matplotlib auto-layout flow; prone to text/shape overlap. Do not use for new reports: author amermaidblock instead.{"t":"code","text":"..."}: monospace block (Consolas on light-gray){"t":"pagebreak"}: force a page break (cover→목차→본문 breaks are automatic)
- Table
statusper row:good= green (pass),bad= red (fail),warn= gray, omitted = neutral. The header row is auto sky-blue and repeats across page breaks.
Follow the type skeleton from Step 1 and the 작성 원칙 above. Language: Korean, plain and direct.
도식 선택 (어떤 블록을 쓸까)
- 구조화된 도식(흐름/워크플로, 시퀀스, 상태, ER, 클래스, 아키텍처) →
mermaid. 노드가 글자에 맞춰 자동으로 커지므로 글자 짤림이 없고, 손으로 좌표를 잡을 필요가 없다. 기본값으로 삼는다. - 정량 비교/추세/비율(막대·선·파이) →
chart. - 표준 그래프로 표현 못 하는 맞춤 그림(특수 기하, 주석 레이아웃, 삽화) → **
svg**를 직접 작성.
Mermaid 도식 작성 규칙 (the mermaid block)
- 문법: 첫 줄에 다이어그램 종류(
flowchart LR|TB,sequenceDiagram,stateDiagram-v2,erDiagram,classDiagram)를 쓰고, 노드/엣지를 이어서 정의한다. 라벨은 한국어로 간결하게. - 방향: 노드가 4개를 넘으면
flowchart LR(가로)보다flowchart TB(세로)가 페이지 폭에 맞아 글자가 더 크게 나온다. 폭이 넘칠 것 같으면 세로로. - 크기:
width_in으로 문서 내 폭을 정한다(기본 6.4). Kroki가 텍스트에 맞춰 렌더하므로 폭만 정하면 된다. - 분기 라벨: 조건 분기는 엣지 라벨(
B -->|성공| C)로 표기한다. - 금지:
%%{init}%%로htmlLabels를 끄지 말 것(불필요). 노드 라벨에—(em-dash) 쓰지 말 것. - Kroki가 각 도식을 서버에서 브라우저로 렌더해 PNG로 굽는다: 글자와 색이 항상 이미지에 박혀 Word/LibreOffice에서 그대로 보인다.
SVG 도식 작성 규칙 (the svg block)
Author the SVG yourself, the way the visualize tool would: deliberate layout, not auto-placed. These rules keep it on-brand and prevent the overlap/distortion that the old matplotlib diagram produced:
- Canvas: set
viewBox="0 0 W H"(this fixes the aspect ratio;width_insizes it in the doc). No pixelwidth/heightneeded. - Font: put
font-family="'맑은 고딕','Malgun Gothic','Apple SD Gothic Neo',sans-serif"on every<text>; titlesfont-weight="bold". - Palette (match the report): text/heading navy
#1F3864; strokes & arrows blue#2E6DA4, good#2EA84F, bad#C0392B, warn#E8862E; light box fills#EAF1F8(blue) /#E2EFDA(good) /#F8D7DA(bad) /#FFF2CC(warn); plain boxes on white. - Boxes: rounded
rx="9",stroke-width="2". Size each box to its text: roughlywidth ≈ 9px × 글자수 + 32,height ≥ 48. Center the label withtext-anchor="middle"and baseline ≈ box-center-y + 5. - Arrows: draw edge-to-edge (start on the source box border, end on the target border: never center-to-center),
stroke-width="2", end with a<marker>arrowhead colored like the line. Put an edge label at the segment midpoint in the matching color. - Spacing: leave ≥ 24px between boxes; never let text touch or overlap a border or another shape.
- Scope: use SVG for schematic diagrams (boxes + arrows + short labels). For quantitative comparison/trend/share, use a
chartblock, not SVG.
Step 3: Generate the .docx
bash ${CLAUDE_SKILL_DIR}/assets/build.sh <topic>.json <output>.docxbuild.sh checks the toolchain and then runs build_report.js, so building is one call.
build_report.js calls figures.py for chart/diagram blocks (matplotlib), renders each mermaid block to PNG via Kroki (curl; override the endpoint with KROKI_URL), and rasterizes each svg block's PNG fallback via LibreOffice (soffice, resolved on PATH → macOS app bundle). Filename convention: CUBRID_<주제>_<유형>_YYYYMMDD.docx.
Step 4: Validate, visually verify, hand off
1) Check and render in one call:
bash ${CLAUDE_SKILL_DIR}/assets/preview.sh <output>.docxIt verifies the .docx structure (zip integrity, required parts, well-formed XML in every part) and then renders every page to PNG, printing the image paths. Read those images. This is a structural check, not full OOXML schema validation.
2) Visual verification: render every page to an image and read them, to catch layout issues the schema can't (clipped chart labels, overlapping text, a note box merging into a table, broken page breaks, color/table problems). This is the PRIMARY defect-catcher; the structure check cannot see any of these. Do not skip it whenever soffice resolves: only skip if LibreOffice is genuinely absent (and then say so explicitly).
Per-diagram check (render → look → fix → repeat): when you read the rendered pages, inspect each figure specifically: is every node label fully inside its box, no text clipped or overlapping, no shape collision, arrows landing on borders, the whole figure within the page width? If a figure is wrong, fix its block (for mermaid: switch LR↔TB, shorten labels, or adjust width_in; for svg: resize the box or canvas) and re-run Step 3 + this render, then look again. Repeat until every diagram is clean. Do not hand off a report with a diagram you have not looked at.
preview.sh resolves LibreOffice itself (PATH, then the macOS app bundle) and tells you what to install if it is missing. After a fix, re-run build.sh then preview.sh: two calls per round.
Note: LibreOffice substitutes 맑은 고딕 if it is not installed locally: the user's Word (with the font) renders correctly; chart text uses AppleGothic baked into the PNGs.
Then ls -la <output>.docx, tell the user the path, and keep the .json (editable source).
Does it work?
Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.
No reviews yet
People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.
- It worked
- Partly
- It didn't
Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.
Your trail
Files you open will collect here, so you can find them again.
Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.
Now do this for your own.
Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.

