소스 정보
- 저장소
- majiayu000/claude-skill-registry
- 최근 소스 활동
- 2026년 6월 23일 12:15
- 감지된 SKILL.md 언어
- 영어
- 스타
- 543
- 포크
- 85
설치 방법
기본적으로 소스를 먼저 확인하는 Prompt가 선택됩니다. 직접 명령으로 전환하거나 로컬 사본을 다운로드할 수도 있습니다.
소스 파일 검토
설치 여부를 결정하기 전에 SKILL.md와 SkillsMP에 표시된 보조 파일을 읽어 보세요.
메뉴
기본적으로 소스를 먼저 확인하는 Prompt가 선택됩니다. 직접 명령으로 전환하거나 로컬 사본을 다운로드할 수도 있습니다.
설치 여부를 결정하기 전에 SKILL.md와 SkillsMP에 표시된 보조 파일을 읽어 보세요.
Codex 또는 Claude로 설치 이 Prompt를 복사해 Codex, Claude 또는 다른 어시스턴트에 붙여 넣으면 Skill 페이지를 검토하고 설치를 진행할 수 있습니다.
직접 명령은 검토 Prompt를 거치지 않습니다. 실행하기 전에 소스를 확인하세요.
npx skills add https://github.com/majiayu000/claude-skill-registry --skill read-paper명령은 한 줄로 유지됩니다. 복사하기 전에 가로로 스크롤해 전체 내용을 확인하세요.
로컬 사본을 원하시나요? SkillsMP에서 현재 제공할 수 있는 파일을 다운로드하세요.
LLM token logprobs and calibration. Per-decision confidence, ECE, Brier, reliability diagrams, low-confidence triage.
Analyze LLM token logprobs and calibration. Use for per-decision confidence, ECE, Brier scores, reliability diagrams, and low-confidence triage.
回顾最近 N 天的 Claude Code 使用记录——扫描原始会话数据,按主题分组汇总"我都做了什么",并从个人操作系统视角输出模式、风险与增删建议。当用户说 /recap、"看看我这几天做了什么"、"回顾一下我最近的会话"、"这两天我用 claude 干了啥"、"活动回顾" 时使用。
SOC 직업 분류 기준
SKILL.md 표시 중
| name | read_paper |
| description | Convert a research paper into structured Obsidian notes and link research projects |
Read a research paper and convert it into:
This skill transforms reading into reusable research ideas.
Reading/Templates/Paper Note.md exactly - preserve ALL headings, emojis, and structureProjects/Projects/)Reading/)[[wiki linking]] required BUT ONLY to projects/papersBEFORE starting, read the template:
view tool to read Templates/Paper Note.mdTemplate location: Templates/Paper Note.md
Rules:
[[Project Name]] - links to files in Projects/[[Paper Title]] - links to influential papers
[[Attention Is All You Need]] not "Attention is all you need"[[BERT: Pre-Training of Deep Bidirectional Transformers for Language Understanding]][[Representation Engineering - A Top-Down Approach to AI Transparency]][[Transformer Architecture]] - this is a concept, NOT allowed[[Attention Mechanism]] - this is a technique, NOT allowed[[Self-Supervised Learning]] - this is a method, NOT allowed[[BERT]] - this is a model name, NOT allowed (unless it's a paper title)[[Cross-Entropy Loss]] - this is a concept, NOT allowedDo NOT link them. Just write them as plain text.
Examples:
[[attention mechanisms]]"[[contrastive learning]]"Exception: Only link if you're creating/referencing a PROJECT about that mechanism:
[[Scaling Laws for Attention Heads]]" (project)When searching for papers online:
https://arxiv.org/html/[paper-id]https://arxiv.org/src/[paper-id]
.tar.gz archive — extract it and read the .tex files.tex file (often main.tex or the largest .tex file)MANDATORY FIRST STEP:
Use view tool on: Templates/Paper Note.md
Study the structure before creating the paper note.
: for -Reading/Templates/Paper Note.mdAnswer each subsection prompt with bullet points:
🌍 Background: Answer both subsection prompts
❓ Problem Statement: Answer all three subsection prompts
💡 Key Idea: Answer all three subsection prompts
🧭 High-Level Approach: Describe method/framework
🔬 Technical Details / Analysis: Answer all three subsection prompts
📊 Main Results / Observations: List findings
🧩 Interpretation: Answer both subsection prompts
⚠️ Limitations / Open Questions: List items
🔗 Connections to Other Work: Fill three subsections with [[Paper Title]] wiki links
[[Title]] not plain textCRITICAL: Write technical details as plain text. Do NOT create wiki links for concepts, methods, or techniques.
Part A: Link to existing projects immediately
Projects/ that relate to the paperPart B: Propose new projects (requires user approval)
CRITICAL: Ground proposals in the paper itself, not inferred extensions
Extract ideas directly from the paper's content:
DO NOT infer extensions beyond what the paper discusses. If the paper doesn't mention it, don't propose it.
Before creating ANY new project files:
AskUserQuestion tool to let the user select which projects to create:
multiSelect: true so the user can pick multiple projectslabel = project title, description = 1-line summary of what it exploresquestion: "Which projects should I create from this paper?"
header: "Projects"
multiSelect: true
options:
- label: "Project Title A", description: "Explores X based on paper's open question about Y"
- label: "Project Title B", description: "Tests whether Z holds, motivated by limitation W"
- label: "None", description: "Don't create any new projects"
Project title examples:
Detecting Representation Drift During Fine-TuningPreventing Emergent Misalignment via Gradient MonitoringFor each approved new project:
Projects/Templates/Project.md as templateseedAfter receiving user approval, create the new project files:
Projects/Templates/Project.md as templateCRITICAL: Bidirectional linking for existing projects
For every project linked from the paper note (both existing and newly created):
Projects/[[Paper Title]] (the paper note file name)This ensures the graph remains consistent: if paper → project link exists, then project → paper link must also exist.
Then update the paper note:
[[Paper Title]] not plain text[[Sparse Autoencoders Find Highly Interpretable Features in Language Models]][[Concept Bottleneck Models]][[sparse autoencoders]] (not title case, too generic)VERIFY: Every [[link]] in the paper note is either:
Projects/ (project note) - appears under "Linked Research Projects"[[Title Case Paper Title]]
Before finishing:
Projects/ or Reading/After completing the paper note (and any new projects):
.claude/vault_maintanance.mdpapers_since_audit_links, papers_since_vault_index, and papers_since_validate_links each by 1projects_since_audit_links and projects_since_vault_index each by the number of new projects created.claude/vault_maintanance.mdProjects/) or papers (Reading/): in file names. Exchange : always with -Convert reading into:
Important workflow:
A paper note without linked projects is considered incomplete. A paper note that doesn't follow the template structure is considered invalid.
Template structure:
Templates/Paper Note.md first? ✅Links:
For EVERY [[link]] in the paper note:
Projects/? ✅Reading/? ✅Bidirectional linking verification:
If uncertain whether something should be a link: Don't link it. Use plain text.
Maintenance counters:
papers_since_audit_links, papers_since_vault_index, papers_since_validate_links in .claude/vault_maintanance.md? ✅projects_since_audit_links and projects_since_vault_index for each new project created? ✅