リサーチ・学習のスキル

Diagnostic guide for RL-based post-training of language models (GRPO, PPO, REINFORCE, DPO). Use proactively when debugging a training pipeline that shows no improvement, loss anomalies, reward stagnation, NaN gradients, or other unexpected behavior during reinforcement learning fine-tuning. Work through all pipeline stages — reward, advantages, log-probs, loss, generation/decoding — before concluding the diagnosis is complete; stopping after one or two fixes is a common failure mode. Covers log-probability math, advantage estimation, numerical stability, reward processing, and generation/decoding pipeline issues.

日本語の概要は準備中です。原文の説明を表示しています。

Comprehensive citation management for academic research. Search Google Scholar and PubMed for papers, extract accurate metadata, validate citations, and generate properly formatted BibTeX entries. This skill should be used when you need to find papers, verify citation information, convert DOIs to BibTeX, or ensure reference accuracy in scientific writing.

日本語の概要は準備中です。原文の説明を表示しています。

This is a summary the advantages and disadvantages of earthquake event detection and phase picking methods, shared by leading seismology researchers at the 2025 Earthquake Catalog Workshop. Use it when you have a seismic phase picking task at hand.

日本語の概要は準備中です。原文の説明を表示しています。

Implement paper-defined attention variants by extracting mechanism invariants, mapping tensor shapes, preserving module interfaces, validating numerical behavior, and integrating the custom attention into an existing transformer stack.

日本語の概要は準備中です。原文の説明を表示しています。

Implement mHC (Manifold-Constrained Hyper-Connections) for stabilizing deep network training. Use when implementing residual connection improvements with doubly stochastic matrices via Sinkhorn-Knopp algorithm. Based on DeepSeek's 2025 paper (arXiv:2512.24880).

日本語の概要は準備中です。原文の説明を表示しています。

Track progress through CSES Problem Set with structured learning

日本語の概要は準備中です。原文の説明を表示しています。

Autonomous research engineering — web search orchestration, source synthesis, hypothesis formation, and structured research report generation.

日本語の概要は準備中です。原文の説明を表示しています。

Complexity-based task routing with Q-Learning optimization, Agent Booster WASM fast-path, and Mixture-of-Experts model selection.

日本語の概要は準備中です。原文の説明を表示しています。

SONA self-optimizing neural architecture with ReasoningBank trajectory learning, EWC++ anti-forgetting, and reinforcement learning feedback loops.

日本語の概要は準備中です。原文の説明を表示しています。

Clarify vague requirements through exploratory questioning and option generation before committing to research or implementation.

日本語の概要は準備中です。原文の説明を表示しています。

Systematic codebase exploration following the Iron Law - understand the problem before exploring code. Four phases with file-finder and web-researcher agents.

日本語の概要は準備中です。原文の説明を表示しています。

Interactive PM interview with expertise-adaptive questioning for requirements elicitation

日本語の概要は準備中です。原文の説明を表示しています。

Fresh adversarial code review with binary PASS/FAIL verdicts, evidence citations, and anchoring bias prevention via fresh reviewer spawning.

日本語の概要は準備中です。原文の説明を表示しています。

Feature specification creation from codebase research. Produces requirements, acceptance criteria, architecture decisions, implementation plans, and risk analysis.

日本語の概要は準備中です。原文の説明を表示しています。

Scaffold new babysitter process definitions following SDK patterns, proper structure, and best practices. Guides the 3-phase workflow from research to implementation.

日本語の概要は準備中です。原文の説明を表示しています。

Structured planning methodology with research, brainstorming, phased plan creation, risk assessment, and plan-to-build continuity.

日本語の概要は準備中です。原文の説明を表示しています。

Research-first development methodology that investigates existing solutions, brainstorms alternatives, and evaluates trade-offs before any implementation begins.

日本語の概要は準備中です。原文の説明を表示しています。

Dynamic context injection, mode switching (dev/review/research), selective loading, and strategic compaction for token optimization.

日本語の概要は準備中です。原文の説明を表示しています。

Parallel research agent orchestration dispatching 5-10 concurrent agents for comprehensive multi-source research with synthesis and validation.

日本語の概要は準備中です。原文の説明を表示しています。

Pattern extraction, confidence-scored evaluation, skill creation, organization, versioning, and cross-project export pipeline.

日本語の概要は準備中です。原文の説明を表示しています。

Self-learning system that captures corrections during sessions and reminds users to run /reflect to update CLAUDE.md. Use when discussing learnings, corrections, or when the user mentions remembering something for future sessions.

日本語の概要は準備中です。原文の説明を表示しています。

Use whenever you generate or review a chart, plot, table, or paper figure in this workspace, including work delegated by paper-writing, literature-survey, and experiment skills. Applies the Open Science publication style, enforces readable final-size layout for figures and tables, and rejects generic diagram-tool output as a publication figure. Interactive Plotly/HTML may be used for exploration, but paper delivery requires a static publication-ready export.

日本語の概要は準備中です。原文の説明を表示しています。

Identifies high-quality leads for your product or service by analyzing your business, searching for target companies, and providing actionable contact strategies. Perfect for sales, business development, and marketing professionals.

日本語の概要は準備中です。原文の説明を表示しています。

Assists in writing high-quality content by conducting research, adding citations, improving hooks, iterating on outlines, and providing real-time feedback on each section. Transforms your writing process from solo effort to collaborative partnership.

日本語の概要は準備中です。原文の説明を表示しています。