Research, academic papers, education, and exam prep (4.2K skills)
CLAUDE.mdを分割ルールファイル(.claude/rules/)で管理するスキル。学習トライアドの書き込み先を分割ファイルにリダイレクトし、CLAUDE.mdには目次だけを生成する（本文は .claude/rules/ を Claude Code が自動で読み込む）。/rules-init, /rules-merge, /rules-add, /rules-status コマンドを提供。CLAUDE.mdが肥大化している場合や、学習トライアドの蓄積を分割管理したい場合に使用。
The English summary is being prepared. Showing the original description.
CLAUDE.mdの「Lessons Learned」セクションを管理するスキル。`/lessons` で学びを追記、`/lessons review` で既存ルールのレビュー・整理を行う。セッション中の修正・失敗・発見を構造的に蓄積し、チーム全体の知見として共有する。「CLAUDE.mdを更新して」「学びを記録して」等の自然言語でも起動可能。
The English summary is being prepared. Showing the original description.
Claude Code の過去のセッション記録（~/.claude の履歴・セッションログ・memory）から、他のメンバーと共有できる運用 Tips を抽出し、公式ドキュメントや論文などの一次情報で裏付けを取って、Markdown・Artifact・Notion に公開する。「過去の会話から Tips をまとめて」「Claude Code の使い方をチームに共有したい」「今月の学びを Tips にして」などの依頼で使う。セッション 1 本の効率採点ではない（それは evaluate-session の領域）。週次のナレッジカード提出は knowledge-cards で行い、このスキルはカードや記録から Tips 文書を作る下流工程にあたる。
The English summary is being prepared. Showing the original description.
Surface new insights, potential trades, and research opportunities from existing vault content. Use when user says "surface", "what am I missing", "find opportunities", or "what should I research next".
The English summary is being prepared. Showing the original description.
Synchronize recent research learnings across all affected theses, sector notes, and macro documents. Use after a research session or when user says "sync", "propagate", or "update everything".
The English summary is being prepared. Showing the original description.
Targeted deep research to fill a specific gap in an existing thesis. Use when user says "deepen", "flesh out", "expand on", "fill in", or specifies a thesis section to improve. Also supports --sync-metrics mode — detects stale financial metrics across every in-scope analytical section of a thesis (not just Key Metrics or Summary) and coherently updates every sentence that depends on them, using a Key-Metrics-table / FMP / web-search waterfall. Use when user says "sync metrics", "update the numbers throughout the note", or "fix stale figures across the whole thesis". The batch form `--sync-metrics --all-flagged` syncs every thesis flagged by the latest `$numbers --all-open` in one operation — detection fanned out to capped READ-ONLY agents, one consolidated confirmation gate, direct sequential writes (agents never write).
The English summary is being prepared. Showing the original description.
Ingest content into the vault as structured research notes. Accepts a URL, a local file path, `--from-brief [date]` (pick stories from an n8n Daily Intel news brief), or no arguments (batch-processes _Inbox/). Use when user shares a URL, says "clip this", "ingest", "ingest from the brief", or "process inbox".
The English summary is being prepared. Showing the original description.
Pull an earnings call transcript from Financial Modeling Prep, extract qualitative signal (management commentary deltas, hedging shifts, Q&A tone), and produce a thesis-delta-first Research note. Use when user says "transcript", "pull earnings call", "transcript diff", or "ingest [TICKER] earnings".
The English summary is being prepared. Showing the original description.
Pull an earnings call (or investor-event) transcript from Quartr (FMP fallback), extract qualitative signal (management commentary deltas, hedging shifts, Q&A tone), and produce a thesis-delta-first Research note. Use when user says "transcript", "pull earnings call", "transcript diff", or "ingest [TICKER] earnings".
The English summary is being prepared. Showing the original description.
Extract a thesis's load-bearing assumptions and scan for internal contradictions — the falsifiable claims the bull case requires, what would falsify each, and where the thesis's own Risks / Bear Case / research already argue against it. Use when user says "assumptions", "what must be true", "load-bearing assumptions", or "does [TICKER] contradict itself".
The English summary is being prepared. Showing the original description.
Generate a backward-looking retrospective of vault activity (addressed callouts + Log entries) over a time window (1w / 1m / 1q), overlay with newsflow + earnings transcripts + price action, and rank trade ideas by the gap between official narrative and market reaction. Output is an immutable Research note. Use when user says "retro", "retrospective", "what did I do this week/month/quarter", "review activity against market", or "recommend trades from recent research".
The English summary is being prepared. Showing the original description.
Targeted deep research to fill a specific gap in an existing thesis. Use when user says "deepen", "flesh out", "expand on", "fill in", or specifies a thesis section to improve. Also supports --sync-metrics mode — detects stale financial metrics across every in-scope analytical section of a thesis (not just Key Metrics or Summary) and coherently updates every sentence that depends on them, using a Key-Metrics-table / FMP / web-search waterfall. Use when user says "sync metrics", "update the numbers throughout the note", or "fix stale figures across the whole thesis". The batch form `--sync-metrics --all-flagged` syncs every thesis flagged by the latest `/numbers --all-open` in one operation — detection fanned out to capped READ-ONLY agents, one consolidated confirmation gate, direct sequential writes (agents never write).
The English summary is being prepared. Showing the original description.
Earnings-season triage across ALL active theses (not just Live Portfolio holdings) — every earnings call reported inside the window gets a Quartr event summary, a Conviction-Trigger/Outstanding-Questions cross-reference, and a coverage check against existing Research notes, producing a ranked deep-read queue for /transcript. Use when user says "earnings sweep", "who reported", "triage earnings", "catch up on earnings season", or after any earnings-heavy stretch. Default window 14 days; `/earnings-sweep 90` covers a full season.
The English summary is being prepared. Showing the original description.
Convert AmeriFlux ERA5 half-hourly meteorological CSV files into EcoSIM hourly climate forcing NetCDF files, compare generated forcing against overlapping measured tower meteorology when a FULLSET file is available, and report large biases or errors. Use when preparing or validating EcoSIM climate inputs for AmeriFlux sites, checking ERA5 against in situ records, diagnosing forcing discrepancies, or writing climate-quality reports. For non-AmeriFlux paper sites that need a fresh Copernicus CDS ERA5 point download by longitude/latitude, use era5-cds-point-download first.
The English summary is being prepared. Showing the original description.
Download Copernicus CDS ERA5 single-level point time-series climate forcing for non-US or non-NLDAS EcoSIM sites from longitude/latitude coordinates. Use when a paper-derived site is outside the NLDAS domain, lacks AmeriFlux/FLUXNET curated forcing, or needs global ERA5 variables such as 2m temperature, dewpoint temperature, surface pressure, solar radiation, precipitation, and 10m wind components for EcoSIM preprocessing.
The English summary is being prepared. Showing the original description.
Extract study sites, observational variables, coordinates, climate-forcing sources, and cited data sources from ecosystem-modeling papers, PDFs, supplements, or article text, then plan or create EcoSIM simulation workflows and run folders for those sites. Use when Codex is asked to read a paper for EcoSIM feasibility, including requests phrased as "read paper for ecosim feasibility"; read a paper for sites/data availability; reproduce a published site simulation in EcoSIM; compare EcoSIM outputs to paper observations; or convert paper-derived AmeriFlux, FLUXNET, US NLDAS-domain, non-US ERA5, crop, peatland, forest, grassland, wetland, or unmanaged field-site information into EcoSIM inputs and runnable cases.
The English summary is being prepared. Showing the original description.
Use this skill when deriving EcoSIM plant trait parameter recommendations for a named plant from web and online literature evidence, using a read-only `plant_trait.*.desc` file and the `ndlf43` tree block or `gr3s43` grass block as the starting archetype. Never edit or replace the `.desc` file.
The English summary is being prepared. Showing the original description.
会話で直近に調べた・生成した内容 (調査結果、比較、設計案、手順など) を Markdown や HTML ファイルに書き出すスキル。ユーザーが「md で出力して」「html にまとめて」「ファイルに書き出して」など会話内容のファイル化の意図を示したら必ず使用する。
The English summary is being prepared. Showing the original description.
research の結果と受け入れ条件をもとに、実装方法を選択し TDD ベースの実装計画と動作確認チェックリストを作成する。
The English summary is being prepared. Showing the original description.
GitHub issue から実装前の調査を行い、受け入れ条件・影響範囲・実装方法の候補・リファレンス・盲点候補を整理する(選択は plan に委ねる)。
The English summary is being prepared. Showing the original description.
Build complex AI systems with declarative programming, optimize prompts automatically, create modular RAG systems and agents with DSPy - Stanford NLP's framework for systematic LM programming
The English summary is being prepared. Showing the original description.
Fine-tune LLMs using reinforcement learning with TRL - SFT for instruction tuning, DPO for preference alignment, PPO/GRPO for reward optimization, and reward model training. Use when need RLHF, align model with preferences, or train from human feedback. Works with HuggingFace Transformers.
The English summary is being prepared. Showing the original description.
Evaluates LLMs across 60+ academic benchmarks (MMLU, HumanEval, GSM8K, TruthfulQA, HellaSwag). Use when benchmarking model quality, comparing models, reporting academic results, or tracking training progress. Industry standard used by EleutherAI, HuggingFace, and major labs. Supports HuggingFace, vLLM, APIs.
The English summary is being prepared. Showing the original description.
Use this skill instead of WebSearch for ANY question requiring web research. Trigger on queries like "what is X", "explain X", "compare X and Y", "research X", or before content generation tasks. Provides systematic multi-angle research methodology instead of single superficial searches. Use this proactively when the user's question needs online information.
The English summary is being prepared. Showing the original description.