Chroma Vector Database
Orchestra-Research/AI-Research-SKILLs
Shows how to store documents and embeddings in Chroma, query them by similarity with metadata filters, and persist them to disk for RAG and semantic search projects.
RAG 파이프라인의 문서 청킹 전략을 체계적으로 설계하는 방법론. An agent skill from revfactory/harness-100.
Install Chunking Strategy Guide
$ npx skills add revfactory/harness-100 --skill chunking-strategy-guide -a claude-codeProject install by default; add -g for ~/.claude/skills/.
$ gh skill install revfactory/harness-100 chunking-strategy-guide --agent claude-codeProject scope by default; add --scope user for a personal install. Needs GitHub CLI 2.90.0 or later (public preview).
$ git clone --depth 1 https://github.com/revfactory/harness-100.git skills-src && mkdir -p .claude/skills && cp -r skills-src/ko/41-llm-app-builder/.claude/skills/chunking-strategy-guide .claude/skills/chunking-strategy-guide && rm -rf skills-srcUse ~/.claude/skills/ instead of .claude/skills for a personal install. The folder must contain SKILL.md.
Claude Code skills documentation · loads skills from .claude/skills/
Install the "chunking-strategy-guide" agent skill from https://github.com/revfactory/harness-100/tree/main/ko/41-llm-app-builder/.claude/skills/chunking-strategy-guide into .claude/skills/chunking-strategy-guide/ in this project. Copy the whole folder (SKILL.md and every file beside it), keep the folder name "chunking-strategy-guide", then confirm the skill loads.Claude Code copies the folder itself, the same result as the manual copy. Check what it changed before you commit it.
$skill-installer install https://github.com/revfactory/harness-100/tree/main/ko/41-llm-app-builder/.claude/skills/chunking-strategy-guideType this inside Codex. $skill-installer <name> installs a curated skill from openai/skills. The installer writes to $CODEX_HOME/skills (default ~/.codex/skills). Restart Codex if the skill does not show up.
$ npx skills add revfactory/harness-100 --skill chunking-strategy-guide -a codexProject install goes to .agents/skills/; add -g for ~/.codex/skills/.
$ gh skill install revfactory/harness-100 chunking-strategy-guide --agent codexProject scope by default (.agents/skills/); add --scope user for a personal install.
$ git clone --depth 1 https://github.com/revfactory/harness-100.git skills-src && mkdir -p .agents/skills && cp -r skills-src/ko/41-llm-app-builder/.claude/skills/chunking-strategy-guide .agents/skills/chunking-strategy-guide && rm -rf skills-srcUse ~/.agents/skills/ instead of .agents/skills for a personal install.
Codex skills documentation · loads skills from .agents/skills/
Install the "chunking-strategy-guide" agent skill from https://github.com/revfactory/harness-100/tree/main/ko/41-llm-app-builder/.claude/skills/chunking-strategy-guide into .agents/skills/chunking-strategy-guide/ in this project. Copy the whole folder (SKILL.md and every file beside it), keep the folder name "chunking-strategy-guide", then confirm the skill loads.Codex copies the folder itself, the same result as the manual copy. Check what it changed before you commit it.
$ npx skills add revfactory/harness-100 --skill chunking-strategy-guide -a cursorProject install goes to .agents/skills/; add -g for ~/.cursor/skills/.
$ gh skill install revfactory/harness-100 chunking-strategy-guide --agent cursorProject scope by default (.agents/skills/); add --scope user for a personal install.
$ git clone --depth 1 https://github.com/revfactory/harness-100.git skills-src && mkdir -p .cursor/skills && cp -r skills-src/ko/41-llm-app-builder/.claude/skills/chunking-strategy-guide .cursor/skills/chunking-strategy-guide && rm -rf skills-srcUse ~/.cursor/skills/ instead of .cursor/skills for a personal install.
Cursor skills documentation · loads skills from .cursor/skills/, .agents/skills/, .claude/skills/, .codex/skills/
Install the "chunking-strategy-guide" agent skill from https://github.com/revfactory/harness-100/tree/main/ko/41-llm-app-builder/.claude/skills/chunking-strategy-guide into .cursor/skills/chunking-strategy-guide/ in this project. Copy the whole folder (SKILL.md and every file beside it), keep the folder name "chunking-strategy-guide", then confirm the skill loads.Cursor copies the folder itself, the same result as the manual copy. Check what it changed before you commit it.
$ gemini skills install https://github.com/revfactory/harness-100.git --path ko/41-llm-app-builder/.claude/skills/chunking-strategy-guide--scope user (default) or --scope workspace; --path is the subfolder of the repo that holds the skill; --consent skips the security confirmation prompt.
$ npx skills add revfactory/harness-100 --skill chunking-strategy-guide -a gemini-cliProject install goes to .agents/skills/; add -g for ~/.gemini/skills/.
$ gh skill install revfactory/harness-100 chunking-strategy-guide --agent gemini-cliProject scope by default (.agents/skills/); add --scope user for a personal install.
$ git clone --depth 1 https://github.com/revfactory/harness-100.git skills-src && mkdir -p .gemini/skills && cp -r skills-src/ko/41-llm-app-builder/.claude/skills/chunking-strategy-guide .gemini/skills/chunking-strategy-guide && rm -rf skills-srcUse ~/.gemini/skills/ instead of .gemini/skills for a personal install, then run /skills reload.
Gemini CLI skills documentation · loads skills from .gemini/skills/, .agents/skills/
Install the "chunking-strategy-guide" agent skill from https://github.com/revfactory/harness-100/tree/main/ko/41-llm-app-builder/.claude/skills/chunking-strategy-guide into .gemini/skills/chunking-strategy-guide/ in this project. Copy the whole folder (SKILL.md and every file beside it), keep the folder name "chunking-strategy-guide", then confirm the skill loads.Gemini CLI copies the folder itself, the same result as the manual copy. Check what it changed before you commit it.
$ gh skill install revfactory/harness-100 chunking-strategy-guideInstalls for Copilot at project scope by default; add --scope user for a personal install. Preview a skill first with gh skill preview. Needs GitHub CLI 2.90.0 or later (public preview).
$ npx skills add revfactory/harness-100 --skill chunking-strategy-guide -a github-copilotProject install goes to .agents/skills/; add -g for ~/.copilot/skills/.
$ git clone --depth 1 https://github.com/revfactory/harness-100.git skills-src && mkdir -p .github/skills && cp -r skills-src/ko/41-llm-app-builder/.claude/skills/chunking-strategy-guide .github/skills/chunking-strategy-guide && rm -rf skills-srcUse ~/.copilot/skills/ instead of .github/skills for a personal install. Commit .github/skills so cloud agent and code review can use it.
GitHub Copilot skills documentation · loads skills from .github/skills/, .claude/skills/, .agents/skills/
Install the "chunking-strategy-guide" agent skill from https://github.com/revfactory/harness-100/tree/main/ko/41-llm-app-builder/.claude/skills/chunking-strategy-guide into .github/skills/chunking-strategy-guide/ in this project. Copy the whole folder (SKILL.md and every file beside it), keep the folder name "chunking-strategy-guide", then confirm the skill loads.GitHub Copilot copies the folder itself, the same result as the manual copy. Check what it changed before you commit it.
$ npx skills add revfactory/harness-100 --skill chunking-strategy-guide -a opencodeOpenCode documents no install command of its own. Project install goes to .agents/skills/; add -g for ~/.config/opencode/skills/.
$ gh skill install revfactory/harness-100 chunking-strategy-guide --agent opencodeProject scope by default (.agents/skills/); add --scope user for a personal install.
$ git clone --depth 1 https://github.com/revfactory/harness-100.git skills-src && mkdir -p .opencode/skills && cp -r skills-src/ko/41-llm-app-builder/.claude/skills/chunking-strategy-guide .opencode/skills/chunking-strategy-guide && rm -rf skills-srcUse ~/.config/opencode/skills/ instead of .opencode/skills for a personal install.
OpenCode skills documentation · loads skills from .opencode/skills/, .claude/skills/, .agents/skills/
Install the "chunking-strategy-guide" agent skill from https://github.com/revfactory/harness-100/tree/main/ko/41-llm-app-builder/.claude/skills/chunking-strategy-guide into .opencode/skills/chunking-strategy-guide/ in this project. Copy the whole folder (SKILL.md and every file beside it), keep the folder name "chunking-strategy-guide", then confirm the skill loads.OpenCode copies the folder itself, the same result as the manual copy. Check what it changed before you commit it.
Facts
- Skill name
chunking-strategy-guide- GitHub stars
- 1.3k
- Token cost
- ~827 tokens
- SKILL.md length
- 205 words
- Files
- 1
- Skills in repo
- 464
- Repo updated
- First seen
- Licence
- Apache-2.0
At a glance
RAG 파이프라인의 문서 청킹 전략을 체계적으로 설계하는 방법론. An agent skill from revfactory/harness-100.
- Tasks that involve Retrieval-augmented generation
- SKILL.md covers 대상 에이전트, 청킹 전략 비교표, 청킹 파라미터 가이드 and 시맨틱 청킹 알고리즘, plus 4 more sections
- Instructions only: no scripts, shell commands, URLs or credentials in SKILL.md
What it does
Chunking Strategy Guide is an agent skill from revfactory/harness-100. RAG 파이프라인의 문서 청킹 전략을 체계적으로 설계하는 방법론. '청킹 전략', '문서 분할', 'RAG 청킹', '임베딩 최적화', '시맨틱 청킹', '텍스트 분할' 등 RAG 데이터 전처리 시 사용한다. 단, 벡터 DB 인프라 구축, 임베딩 모델 학습은 이 스킬의 범위가 아니다.
Its SKILL.md is about 830 tokens, which your agent loads only when the skill is triggered. It is a single SKILL.md file with no bundled scripts.
It sits in AI & LLM Engineering, covering Retrieval-augmented generation. The licence is Apache-2.0.
When your agent uses it
- Tasks that involve Retrieval-augmented generation
Example prompts
- “RAG 청킹”
- “/chunking-strategy-guide”
Requirements
- Python 3
What it can do on your machine
Read from SKILL.md and the folder at commit 8e8d35c. It shows what the files ask for, not the result of running them.
Tool permissions
Pre-approves nothing: there is no allowed-tools line, so your agent's usual permission prompts apply.
From allowed-tools in the SKILL.md frontmatter.
Runs code
No scripts in the folder and no shell commands in SKILL.md (its code samples are python).
From the folder's file list and the shell code blocks in SKILL.md.
Network
No URLs in SKILL.md.
From URLs in SKILL.md, links to its own repository left out.
Credentials
Names no API keys, tokens, secrets or passwords.
From names ending in _API_KEY, _TOKEN, _SECRET, _KEY or _PASSWORD in SKILL.md.
Context cost
Chunking Strategy Guide loads about 827 tokens when it runs. Until then it costs about 46 tokens; SKILL.md has 205 words of instructions outside code blocks.
Estimates: characters ÷ 4, the usual rule of thumb; real counts depend on the model's tokenizer. Scripts and assets cost tokens only if the agent reads them.
Safety
Auto-check passed
The automated check found no risky patterns in SKILL.md.
Automated static check — not a guarantee. Review scripts before installing. It scans the text of SKILL.md for risky patterns (piping downloads into a shell, reading credential files, hidden Unicode, destructive commands); files beside SKILL.md are not scanned.
SKILL.md
The full file from revfactory/harness-100 at commit 8e8d35c, republished under its Apache-2.0 licence (© revfactory). 205 words, ~827 tokens.
Download SKILL.mdSave it as
.claude/skills/chunking-strategy-guide/SKILL.md (or your agent's skills folder).- name
- chunking-strategy-guide
- description
- RAG 파이프라인의 문서 청킹 전략을 체계적으로 설계하는 방법론. '청킹 전략', '문서 분할', 'RAG 청킹', '임베딩 최적화', '시맨틱 청킹', '텍스트 분할' 등 RAG 데이터 전처리 시 사용한다. 단, 벡터 DB 인프라 구축, 임베딩 모델 학습은 이 스킬의 범위가 아니다.
Chunking Strategy Guide — RAG 문서 청킹 전략
rag-architect의 데이터 전처리 역량을 강화하는 스킬.
대상 에이전트
- rag-architect — 문서를 효과적으로 청킹하여 검색 품질을 높인다
- eval-specialist — 청킹 전략의 검색 품질을 평가한다
청킹 전략 비교표
| 전략 | 원리 | 장점 | 단점 | 적합 |
|---|---|---|---|---|
| 고정 크기 | N토큰 단위 절단 | 구현 간단 | 의미 단절 | 로그, 코드 |
| 문장 기반 | 문장 단위 분리 | 의미 보존 | 크기 불균등 | 뉴스, 블로그 |
| 단락 기반 | 빈 줄 기준 | 논리 단위 유지 | 단락 크기 편차 | 문서, 리포트 |
| 시맨틱 | 임베딩 유사도 기준 | 최고 품질 | 느림, 비용 | 복잡한 문서 |
| 재귀적 | 계층적 분리자 | 균형적 | 설정 복잡 | 범용 |
| 마크다운 | 헤딩 기준 | 구조 보존 | MD 전용 | 기술 문서 |
청킹 파라미터 가이드
최적 청크 크기
| 문서 유형 | 청크 크기 | 오버랩 | 이유 |
|----------|----------|--------|------|
| FAQ | 100-200 토큰 | 0 | 질문-답변 쌍이 짧음 |
| 기술 문서 | 300-500 토큰 | 50 | 코드+설명 단위 |
| 법률 문서 | 500-800 토큰 | 100 | 조항 단위 |
| 학술 논문 | 400-600 토큰 | 80 | 단락 단위 |
| 채팅 로그 | 200-300 토큰 | 30 | 대화 턴 단위 |
| 소설/에세이 | 300-500 토큰 | 50 | 장면/단락 단위 |오버랩 비율 공식
optimal_overlap = chunk_size * 0.1 ~ 0.2
규칙:
- 독립적 문서 (FAQ): 오버랩 0
- 연속적 문서 (매뉴얼): 10-15%
- 고밀도 문서 (법률): 15-20%
- 최대 오버랩: chunk_size의 25% 초과 금지시맨틱 청킹 알고리즘
def semantic_chunking(text, model, threshold=0.5):
"""
1. 문장 단위로 분리
2. 인접 문장 쌍의 임베딩 코사인 유사도 계산
3. 유사도가 threshold 이하인 지점에서 분리
4. 최소/최대 청크 크기 제약 적용
"""
sentences = split_sentences(text)
embeddings = model.encode(sentences)
breakpoints = []
for i in range(len(embeddings) - 1):
sim = cosine_similarity(embeddings[i], embeddings[i+1])
if sim < threshold:
breakpoints.append(i + 1)
chunks = split_at(sentences, breakpoints)
return enforce_size_limits(chunks, min=100, max=800)문서 유형별 전처리 파이프라인
PDF → 텍스트 추출 (pdfplumber/pymupdf)
→ 헤더/푸터 제거
→ 페이지 번호 제거
→ 표 → 마크다운 변환
→ 이미지 → alt text / OCR
→ 메타데이터 추출 (제목, 저자, 날짜)
→ 청킹HTML/웹페이지
HTML → 본문 추출 (trafilatura/readability)
→ 네비게이션/사이드바/광고 제거
→ 마크다운 변환
→ 링크 텍스트 보존 ([텍스트](URL))
→ 테이블 보존
→ 메타데이터 추출 (title, description)
→ 청킹코드
코드 → AST 파싱
→ 함수/클래스 단위 분리
→ docstring + 시그니처 + 본문
→ 파일 경로 메타데이터 추가
→ 관련 테스트 코드 연결
→ 청킹 (함수 단위)메타데이터 강화 전략
chunk_with_metadata = {
"text": "청크 텍스트...",
"metadata": {
"source": "document.pdf",
"page": 5,
"section": "3.2 아키텍처",
"heading_hierarchy": ["3. 설계", "3.2 아키텍처"],
"chunk_index": 12,
"total_chunks": 45,
"created_at": "2025-01-15",
"document_type": "technical_spec",
"language": "ko"
}
}청킹 품질 평가 메트릭
| 메트릭 | 공식 | 기준 |
|---|---|---|
| 정보 완전성 | 원본 핵심 정보 / 전체 핵심 정보 | >= 95% |
| 의미 단절률 | 문장 중간 절단 / 전체 청크 | <= 5% |
| 크기 균일성 | 1 - (std / mean) | >= 0.7 |
| 검색 정밀도 | 관련 청크 / 반환 청크 (top-5) | >= 60% |
| 검색 재현율 | 반환 관련 / 전체 관련 (top-10) | >= 80% |
임베딩 모델 선택 가이드
| 모델 | 차원 | 한국어 | 비용 | 용도 |
|---|---|---|---|---|
| text-embedding-3-small | 1536 | 양호 | $0.02/1M | 범용, 비용 효율 |
| text-embedding-3-large | 3072 | 양호 | $0.13/1M | 고품질 |
| multilingual-e5-large | 1024 | 우수 | 무료(로컬) | 한국어 특화 |
| bge-m3 | 1024 | 우수 | 무료(로컬) | 다국어, 긴 컨텍스트 |
| voyage-multilingual-2 | 1024 | 우수 | $0.12/1M | 다국어 최고 |
© revfactory, Apache-2.0. Rendered from Markdown: HTML in the file is shown as text, images as links, and headings moved down two levels. Raw file
Files
Just SKILL.md in ko/41-llm-app-builder/.claude/skills/chunking-strategy-guide of revfactory/harness-100.
Open the folder on GitHubat commit 8e8d35c
Compare with similar skills
Chunking Strategy Guide next to the 5 skills that share the most tags, products or categories with it. Stars are the repository's; “used in” counts other GitHub owners with a copy.
| Skill | Stars | Used in | Tokens | Auto-check | Licence | Repo updated |
|---|---|---|---|---|---|---|
| Chunking Strategy Guide this skillrevfactory/harness-100 | 1.3k | — | ~827 | Automated safety check: Pass | Apache-2.0 | |
| Chroma Vector DatabaseOrchestra-Research/AI-Research-SKILLs | 13k | 7 repos | ~2.3k | Automated safety check: Pass | MIT | |
| Senior Prompt Engineermaslennikov-ig/claude-code-orchestrator-kit | 260 | 3 repos | ~1.4k | Automated safety check: Pass | Custom licence | |
| MCP Local RAGshinpr/mcp-local-rag | 411 | — | ~4.4k | Automated safety check: Pass | MIT | |
| Ms Agent Framework RAGshuyu-labs/WebCode | 278 | — | ~1.1k | Automated safety check: Pass | Custom licence | |
| Local RAG Searchnkapila6/mcp-local-rag | 134 | 1 repos | ~1.6k | Automated safety check: Pass | MIT |
Similar skills
Senior Prompt Engineer
maslennikov-ig/claude-code-orchestrator-kit
Provides reference guides and Python scripts for prompt optimization, RAG evaluation, and agent orchestration when building or tuning LLM systems.
MCP Local RAG
shinpr/mcp-local-rag
Searches, saves, and maintains a local document index through a local RAG MCP server.
Ms Agent Framework RAG
shuyu-labs/WebCode
Comprehensive guide for building Agentic RAG systems using Microsoft Agent Framework in C.
Local RAG Search
nkapila6/mcp-local-rag
Efficiently perform web searches using the mcp-local-rag server with semantic similarity ranking.
Blockify Integration
iternal-technologies-partners/blockify-agentic-data-optimization
Process documents with Blockify API to create optimized IdeaBlocks for RAG.
More from revfactory/harness-100
All 464 skills in this repoAnti Bot Analyzer
revfactory/harness-100
A skill for analyzing website anti-bot defense mechanisms and developing legitimate evasion strategies.
API Error Design Patterns
revfactory/harness-100
Reference for designing how an API reports failures: structured error codes, response shapes, client-friendly messages, an error catalog and retry or fallback advice.
API Security Checklist
revfactory/harness-100
Walks a backend-dev agent through OWASP API Top 10 checks, authentication and authorization patterns, and defense code during API design.
Arg Parser Generator
revfactory/harness-100
Methodology for systematically designing and generating CLI tool argument parser structures.
Audience Segmentation
revfactory/harness-100
Audience segmentation skill used by the analyst and curator agents.
Audio Storytelling
revfactory/harness-100
Audio storytelling skill used by the podcast scriptwriter and show note editor.
Related
Categories
Questions about Chunking Strategy Guide
What does Chunking Strategy Guide do?
RAG 파이프라인의 문서 청킹 전략을 체계적으로 설계하는 방법론. An agent skill from revfactory/harness-100. Chunking Strategy Guide is an agent skill from revfactory/harness-100. RAG 파이프라인의 문서 청킹 전략을 체계적으로 설계하는 방법론.
When should I use Chunking Strategy Guide?
Chunking Strategy Guide fits situations like: tasks that involve Retrieval-augmented generation.
How do I install Chunking Strategy Guide in Claude Code?
Run `npx skills add revfactory/harness-100 --skill chunking-strategy-guide -a claude-code`. Or copy the skill folder (ko/41-llm-app-builder/.claude/skills/chunking-strategy-guide in revfactory/harness-100) into .claude/skills/chunking-strategy-guide in your project. Claude Code loads it when a task matches its description.
How do I install Chunking Strategy Guide in Codex?
Run `npx skills add revfactory/harness-100 --skill chunking-strategy-guide -a codex`. Or copy the skill folder (ko/41-llm-app-builder/.claude/skills/chunking-strategy-guide in revfactory/harness-100) into .agents/skills/chunking-strategy-guide in your project. Codex loads it when a task matches its description.
Can I use Chunking Strategy Guide in Cursor, Gemini CLI or GitHub Copilot?
Cursor, Gemini CLI, GitHub Copilot and OpenCode also load SKILL.md folders. With the skills CLI, run `npx skills add revfactory/harness-100 --skill chunking-strategy-guide -a cursor` (or -a gemini-cli, github-copilot or opencode for the others). To copy it by hand, put the folder in .cursor/skills/chunking-strategy-guide, .gemini/skills/chunking-strategy-guide, .github/skills/chunking-strategy-guide and .opencode/skills/chunking-strategy-guide in your project.
What does Chunking Strategy Guide need to run?
SKILL.md names no scripts, command-line tools or credentials: Chunking Strategy Guide is instructions for the agent only. Our summary lists: Python 3.
Does Chunking Strategy Guide access the network?
SKILL.md contains no URLs. Any network use would come from the scripts or tools the agent runs. This is read from the text; nothing was executed.
Is Chunking Strategy Guide safe to install?
Our automated static check of SKILL.md found no risky patterns, such as piping downloads into a shell, reading credential files or hidden Unicode. It is not a guarantee. Review the folder before installing.
What licence does Chunking Strategy Guide use?
Chunking Strategy Guide is published under the Apache-2.0 licence (the repository's licence). It allows redistribution, so the full SKILL.md is shown on this page.
How many tokens does Chunking Strategy Guide use?
About 827 tokens (SKILL.md is roughly 3.3k characters). Agents keep only the skill's name and description in context until a task matches; then they load SKILL.md in full.
What are the alternatives to Chunking Strategy Guide?
Skills that share tags, products or a category with Chunking Strategy Guide: Chroma Vector Database (Orchestra-Research/AI-Research-SKILLs, 13k stars), Senior Prompt Engineer (maslennikov-ig/claude-code-orchestrator-kit, 260 stars), MCP Local RAG (shinpr/mcp-local-rag, 411 stars) and Ms Agent Framework RAG (shuyu-labs/WebCode, 278 stars). The comparison table on this page puts their stars, adoption, token cost, safety result and licence side by side.
Who maintains Chunking Strategy Guide?
revfactory (a GitHub user) maintains it in revfactory/harness-100, which has 1,295 GitHub stars. The repository holds 464 skills in this directory. The repository was last updated on March 22, 2026.
Source: revfactory/harness-100 on GitHub. Facts on this page come from the repository at the commit we read; the author's words are quoted as theirs.