Getting it into your agent
It runs from inside its repository, so the clone comes first — what it calls does not travel with the file alone.
git clone --depth 1 https://github.com/lsj4232/KR_PATENT_SKILLnpx agentmods add skills/lsj4232/kr_patent_skill/patent-block-diagram-refineWrote 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/skills/lsj4232/kr_patent_skill/patent-block-diagram-refine)<a href="https://agentmods.dev/skills/lsj4232/kr_patent_skill/patent-block-diagram-refine"><img src="https://agentmods.dev/badge/skills/lsj4232/kr_patent_skill/patent-block-diagram-refine/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/skills/lsj4232/kr_patent_skill/patent-block-diagram-refine"><img src="https://agentmods.dev/badge/skills/lsj4232/kr_patent_skill/patent-block-diagram-refine.svg" alt="Reviewed on agentmods" width="80" height="20"></a>- NVIDIA SkillSpector pass
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 | $0.00358 | $0.02403 |
| Opus 5 | $0.00179 | $0.01202 |
| Sonnet 5 | $0.00072 | $0.00481 |
| Haiku 4.5 | $0.00036 | $0.00240 |
Grade A, and why
patent-block-diagram-refine 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 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.
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 — 132 lines — stays where its author put it; the contents beside it link to each section on GitHub.
특허 도면 정제 (.pbd 수정)
무엇을 하는가
완성된 .pbd(에디터 네이티브 JSON)를 입력으로 받아, 기존 스타일을 유지한 채 부분 수정한다.
새로 그리는 게 아니라 고치는 작업이므로, 착수 전에 입력 파일의 현재 규약을 먼저 읽어
"이 도면이 지금 어떤 fs·shape·부호·간격을 쓰는지"를 스냅샷으로 확정하고, 수정은 그 스냅샷을
복제하는 방향으로만 한다. 임의로 새 값을 도입하지 않는다.
철칙: 렌더 없이 좌표만 믿지 않는다. 수정 → CDP 포트로 렌더 → PNG를 Read로 육안 확인 → 보정. 최소 1회전, 실전에서 3회전까지 돈다.
입력
- 단일
.pbd또는 도면 세트 폴더(도면1.pbd…도면N.pbd). - (있으면) 사건 폴더의
도면_스타일가이드.md— 부호 체계 SINGLE SOURCE. 없으면 pbd에서 추출. - (권장) 명세서 최신
.docx경로 — 부호·단계명 정합 대조용.
앱·렌더 (patent-block-diagram과 공유)
- 앱:
$APP_DIR(Electron, 레포 동봉 소스 =patent-block-diagram-app/) - 렌더 스크립트:
~/.claude/skills/patent-block-diagram/scripts/render_pbd.mjs - 작업용 실행(CDP):
cd "$APP_DIR" && \
./node_modules/.bin/electron.cmd . --remote-debugging-port=9222 \
--user-data-dir="$TMP/pbd-agent" > /dev/null 2>&1 &
# 9초쯤 기다린 뒤
node "~/.claude/skills/patent-block-diagram/scripts/render_pbd.mjs" <입력.pbd> <출력.png> [scale]
⚠️ --user-data-dir 없이 띄우거나 taskkill 로 electron 을 싹 죽이면 사용자가 편집 중인
창을 빼앗는다(단일 인스턴스 락이 userData 폴더 단위). 절대 하지 않는다.
scale 2 = 검토용, 3 = 출원용. .pbd 스키마·화살표 라우팅 규칙은 patent-block-diagram
SKILL.md가 단일 진실(중복 서술 안 함). 여기서는 수정 워크플로우만 규정한다.
작업 순서
1. 스타일 스냅샷 추출 (수정 전 필수)
입력 .pbd(들)를 파싱해 아래를 표로 뽑는다. 이게 이번 수정의 준수 기준이 된다.
python3 - <<'PY'
import json,glob,os
for f in sorted(glob.glob('*.pbd')):
d=json.load(open(f,encoding='utf-8'))
b=d.get('boxes',[]); a=d.get('arrows',[]); t=d.get('tags',[])
fss=sorted({x.get('fs') for x in b})
sh={}
for x in b: sh[x['shape']]=sh.get(x['shape'],0)+1
cont=[x['id'] for x in b if x.get('textTop') or x['w']>1000]
print(f"{f}: box={len(b)} arr={len(a)} tag={len(t)} fs={fss} shapes={sh} nextId={d.get('nextId')} containers={cont}")
print(" tags:", [x['label'] for x in t])
PY
여기서 확정: 이 도면의 표준 fs, 도형 용법, 부호 부여 규칙(S계열/장치/개념), nextId(새 id는 이보다 커야 함).
2. 하우스 스타일 대조
스냅샷을 references/도면_스타일가이드_템플릿.md §1과 대조. 벗어난 값이 있으면
"기존 도면이 이미 비표준"인 것이므로, 기존 도면 값을 우선한다(정합이 표준 준수보다 위).
새로 추가하는 요소만 §1 규약을 따른다.
What ships with it
1 file beside SKILL.md in the same directory: the scripts, references and assets a skill reads on demand. Not counted in the per-session cost; read them before you install if any of them is executable.
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.
- 12d ago First seen · 132 lines · 358 tokens per session scan A f7f6d25b36e0
patent-block-diagram-refine is a skill published in the GitHub repository lsj4232/KR_PATENT_SKILL (15 stars, last pushed 19d ago), licensed MIT. It adds 358 tokens to every session and 2,403 once invoked, about $0.0018 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.
Other skills, from other repositories
eac-habre
Chambres Africaines Extraordinaires (CAE) / Extraordinary African Chambers in the Courts of Senegal (EAC) research, drafting, and analysis — the Hissène Habré case. Use whenever the conversation involves the CAE/EAC, the AU-Senegal Accord of 22 August 2012, the Statute of the Chambers, the Habré trial (2015-2016), the…
jep
Jurisdicción Especial para la Paz (JEP) — Colombia's Special Jurisdiction for Peace — research, drafting, and analysis. Use whenever the conversation involves the JEP, the SIVJRNR / Sistema Integral para la Paz, the 2016 Final Peace Agreement with the FARC-EP, Acto Legislativo 01 de 2017, Ley Estatutaria 1957 de 2019…
nuremberg-tokyo
Post-WWII international military tribunals research, drafting, and analysis: the IMT at Nuremberg (1945-46), the twelve NMT trials under Control Council Law No. 10 (1946-49), and the IMTFE at Tokyo (1946-48). Use whenever the conversation involves Nuremberg, the Tokyo Trial, the London or Tokyo Charters, the IMT or…
reg-64-kosovo
Regulation 64 Panels (Kosovo) research, drafting, and analysis — the UNMIK international judges and prosecutors (IJP) mechanism in the Kosovo district courts, 2000-2008/09. Use whenever the conversation involves UNMIK Regulation 2000/64, the 64 Panels, Regulations 2000/6 and 2001/2, UNSC Resolution 1244 (1999), the…
special-panels-timor-leste
Special Panels for Serious Crimes of the Dili District Court (SPSC, 2000-2006) research, drafting, and analysis. Use whenever the conversation involves the Special Panels, the Serious Crimes Unit (SCU), UNTAET and its Regulations 1999/1, 2000/11, 2000/15 and 2000/30, the 1999 referendum violence, the hybrid…
cps-rca
Cour Pénale Spéciale de la République Centrafricaine (CPS) / Special Criminal Court in the Central African Republic (SCC-CAR) research, drafting, and analysis. Use whenever the conversation involves the CPS, the Loi organique n°15.003 of 3 June 2015, the RPP (Loi 18.010), the affaire Paoua / Lemouna et Koundjili (Issa…