kordoc is a command-line tool and MCP server that reads Korean document formats and converts, compares, analyzes, fills, or generates them, including HWP, PDF, Office files, and images. Government workers and other users handling Korean documents use it to make their contents available as Markdown and automate document processing, while the catalogue entries provide integrations and workflows for coding agents.
Borrowing it
Nothing to install: this file belongs to chrisryugj/kordoc. Take a copy, put it at the same path in your own repository, and replace the rules that are about this project with yours.
curl -O https://raw.githubusercontent.com/chrisryugj/kordoc/main/CLAUDE.mdgit clone --depth 1 https://github.com/chrisryugj/kordocWrote 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/instructions/chrisryugj/kordoc/claude-md)<a href="https://agentmods.dev/instructions/chrisryugj/kordoc/claude-md"><img src="https://agentmods.dev/badge/instructions/chrisryugj/kordoc/claude-md.svg" alt="Measured on agentmods" height="20"></a>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 | $0.09071 | $0.09071 |
| Opus 5 | $0.04535 | $0.04535 |
| Sonnet 5 | $0.01814 | $0.01814 |
| Haiku 4.5 | $0.00907 | $0.00907 |
Grade A, and why
kordoc CLAUDE.md 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 today.
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 — 231 lines — stays where its author put it; the contents beside it link to each section on GitHub.
CLAUDE.md
This file provides guidance to Claude Code (claude.ai/code) when working with code in this repository.
프로젝트 개요
kordoc — 한국 공문서(HWP 5.x, HWPX, PDF, XLSX, DOCX)를 마크다운으로 변환하는 파서 라이브러리.
npm 패키지로 배포되며, 3가지 인터페이스 제공: 라이브러리 API, CLI(kordoc), MCP 서버(kordoc-mcp).
빌드 & 개발
npm run build # tsup으로 ESM+CJS 듀얼 빌드 → dist/
npm run dev # watch 모드
npm test # node --test + tsx 로더 (tests/*.test.ts)
npm run bench:gate # 코퍼스 회귀 게이트 체인 (score·roundtrip·pdf-table·formats·fuzz·reflow·pages) — prepublishOnly에 배선
npm run bench:visual # 한컴 실렌더 시각 오라클 (macOS GUI 전용, 발행 전 수동 1회 — bench/visual/, 순수 로직은 hash-lib.mjs)
ESLint, Prettier 미설정 상태. 벤치 코퍼스(bench/corpus/)는 gitignore — 없으면 맥미니(ssh sm)에서 rsync.
코퍼스 동기화 (2026-09-05 기준선)
bench/corpus/ 아래는 전부 파서 게이트 모수다(recall 1 강제). 생성 엔진 학습용으로 긁은 실결재 원본은
bench/corpus-gen/(gitignore, 2026-09-06 opengov-2609/ 280건)에 두고 corpus 에 섞지 말 것 — 섞으면 파서가 못 읽는
문서 1건에 게이트가 죽는다(36900720 recall 미달로 실측). corpus-gen 도 맥미니와 rsync 로 맞춘다.
맥북·맥미니 양쪽 bench/corpus/ 는 바이트 동일로 맞춰 둔다. 2026-09-05 에 법령 별지서식
licbyl/(법제처 licbyl API 표본 300건 — HWP5 원본 + PDF + rhwp v0.8.6 export-hwpx 변환 HWPX,
bench/collect-licbyl.mjs seed 20260905 로 재현) 900파일, 같은 날 v4.12.2 에서 licbyl2/(서식 2차,
seed 20260906 --exclude=licbyl, 279쌍 — 표본 풀 소진으로 300 미달) 837파일과 licbyl-byl/(별표
--knd=1, 90쌍) 271파일, v4.12.3 에서 licbyl-byl2/(별표 3차, seed 20260907 --knd=1 --exclude=licbyl,licbyl2,licbyl-byl, 183쌍 — 풀 소진으로 200 미달) 549파일이 들어와 게이트 모수는 hwpx ~1,390·pdf
~950·hwp쌍 ~860 이다. rhwp 변환본 가운데 자기참조 GT 정렬이 깨지는 것(licbyl 5 + licbyl2 11 + licbyl-byl2 1)은
known-false-miss/ 에 격리했다(README
참조 — 판정 기준은 HWP5 쌍 유사도 1 + 미스 문자열이 파서 출력에 있음, 둘 다 확인하고 옮길 것).
rhwp 바이너리는 GH release v0.8.6 macos-aarch64(스크래치에 두고 rhwp export-hwpx in.hwp out.hwpx).
종전(2026-08-22) 모수는 hwpx 350·pdf 92·hwp쌍 23 — score.mjs 의 MIN_POP 하한(170/25/12)에 여유가 있다.
HWP5↔PDF 셀 대조 보고 지표는 node bench/cmp-hwp-pdf.mjs licbyl [--linebreaks](licbyl 0.965·licbyl2
0.966·별표 0.746·별표 3차 0.787 — 별표가 낮은 원인은 v4.12.3 실측으로 A 1×1 틀 PDF 미감지 8 / B HWP5
flattenLayoutTables 해체 vs PDF 1열 유지 15 / C 구조 차 7(수식 셀·쪽 경계 분할, 파서 결함 아님) — 게이트 아님).
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.
- today Changed · +48 lines · +3,461 tokens per session c05b21cfaa5f
- 8d ago First seen · 183 lines · 5,610 tokens per session scan A 78460f3ca7f2
kordoc CLAUDE.md is an instructions file published in the GitHub repository chrisryugj/kordoc (1,797 stars, last pushed yesterday), licensed MIT. It adds 9,071 tokens to every session, about $0.0454 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 instructions, from other repositories
codex AGENTS.md
AGENTS.md instructions for openai/codex, covering rust/codex-rs, the codex-core crate, code review rules, crate api surface and model visible context.
vscode buildNext.instructions.md
Working notes and architecture documentation for the new esbuild-based build system in build/next. Use when making changes to the new build pipeline (transpile/bundle commands, NLS plugin, source-map handling, resource copying, or self-hosting watch tasks).
next.js AGENTS.md
AGENTS.md instructions for vercel/next.js, covering next.js development guide, codebase structure, monorepo overview, core package: packages/next and other important packages.
langchain AGENTS.md
AGENTS.md instructions for langchain-ai/langchain, covering global development guidelines for the langchain monorepo, corridor security analysis, project architecture and context, monorepo structure and development tools & commands.
vscode oss-third-party-notices.instructions.md
Instructions for microsoft/vscode, covering vs code oss third-party-notices pipeline, architecture, pipeline flow in ci, applying the notice (cutover) and fallback chain (never fail the build).
spec-kit AGENTS.md
AGENTS.md instructions for github/spec-kit, covering agents.md, about spec kit and specify, quickstart — add a new integration in 5 steps, integration architecture and integrationmanifest — file tracking.