Skills（SKILL.md）は、AIエージェント（Claude Code、Cursor、Codexなど）に特定の能力を追加するための設定ファイルです。
詳しく見る →Prepare for journalism interviews with research checklists, question frameworks, and attribution guidelines. Use when preparing to interview sources, planning follow-up questions, or managing interview logistics. Covers consent, recording laws, and professional protocols.
Interview management, transcription workflows, and source note-taking for journalists. Use when preparing for interviews, managing recordings, transcribing audio/video, organizing source notes, creating timestamped references, or building interview databases. Essential for reporters conducting interviews and managing source relationships.
TwigComponents that re-render dynamically via AJAX. Build reactive UIs in PHP + Twig with zero JavaScript. Every user interaction triggers a server round-trip that re-renders the component and morphs
Modest JavaScript framework that connects JS objects to HTML via data attributes. Stimulus does not render HTML -- it augments server-rendered HTML with behavior.
Modern frontend stack for Symfony. Build reactive UIs with minimal JavaScript using server-rendered HTML.
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
Use this skill whenever the user wants to install, configure, or use asdf (asdf-vm), the universal version manager. Trigger for any mention of asdf, .tool-versions files, managing runtime versions, switching between versions of Node.js, Python, Ruby, Go, Terraform, kubectl, Java, Erlang, Elixir, or any other tool managed by asdf. Also trigger when migrating from nvm, pyenv, rbenv, goenv, tfenv, or similar single-language version managers. Use this skill for help with asdf plugins, asdf install, asdf set/global/local, troubleshooting shims, Fish/Bash/Zsh shell configuration, and multi-project version isolation workflows.
>
Category and brand ungating — requirements, documentation, appeal process, restricted categories guide
Product compliance and safety — certifications, labeling requirements, restricted substances, documentation
Deep learning cell and nucleus segmentation from fluorescence and brightfield microscopy images. Uses pre-trained models (cyto3, nuclei, tissuenet) and a generalist flow-based algorithm that segments cells without requiring retraining on new image types. Outputs label masks for downstream morphology measurement and tracking. Use scikit-image watershed for rule-based segmentation; use Cellpose when deep learning generalization across staining conditions is needed.
Query CELLxGENE Census (61M+ cells) programmatically. Search by cell type, tissue, disease, organism. Get expression matrices as AnnData, stream large queries out-of-core, train PyTorch models on single-cell data. For analyzing your own data use scanpy; for annotated data manipulation use anndata.
Batch correction for single-cell RNA-seq (and other omics) with Harmony. Removes technical batch effects from PCA embeddings while preserving biological variation. Use after PCA, before UMAP/neighbors. Fast and scalable to millions of cells. Python (harmonypy, scanpy integration) and R (Seurat) APIs.
Benchling R&D platform Python SDK. CRUD operations on registry entities (DNA, RNA, proteins, custom entities), inventory management (containers, boxes, transfers), electronic lab notebook entries, workflow automation. Requires Benchling account and API key. For local sequence analysis use biopython; for chemical databases use pubchem.
Opentrons Protocol API v2 for OT-2 and Flex liquid handling robots. Write Python protocols for automated pipetting, serial dilutions, PCR setup, plate replication. Control hardware modules (thermocycler, heater-shaker, magnetic, temperature). For multi-vendor lab automation use pylabrobot.
Adaptyv Bio provides an API and Python SDK for ordering and managing cell-free protein expression and binding assay experiments. Use it to submit protein sequences for expression (scale 10–100 µg), characterize binding affinity (KD) against target proteins, track experiment status, and retrieve results — all programmatically without wet-lab setup. Designed for protein engineers, ML-guided directed evolution, and antibody/nanobody optimization workflows. Requires an Adaptyv Bio account and API key.
>-
Process-based discrete-event simulation framework. Model systems with queues, shared resources, and time-based events: manufacturing lines, service operations, network traffic, logistics. Processes are Python generators that yield events. Resource types include capacity-limited (Resource, PriorityResource, PreemptiveResource), bulk material (Container), and object storage (Store, FilterStore). For continuous simulation use SciPy ODE solvers; for agent-based modeling use Mesa.