Skills
96,981 packaged capabilities and playbooks for agents. Install any of them with one command.
Data & databases
Use when the user is doing AI/ML work in a scientific domain such as biology, chemistry, physics, astronomy, climate, genomics, materials, medicine, ecology, energy, engineering, math, drug discovery, protein design, weather modeling, theorem proving, single-cell, or PDE solving. Hugging Science is a curated catalog of scientific datasets, models, blog posts, and interactive Spaces. This skill helps discover and use resources via `datasets`, `transformers`, the HF Inference API, `gradio_client`, and methodology citations.
Use when the user is doing AI/ML work in a scientific domain such as biology, chemistry, physics, astronomy, climate, genomics, materials, medicine, ecology, energy, engineering, math, drug discovery, protein design, weather modeling, theorem proving, single-cell, or PDE solving. Hugging Science is a curated catalog of scientific datasets, models, blog posts, and interactive Spaces. This skill helps discover and use resources via `datasets`, `transformers`, the HF Inference API, `gradio_client`, and methodology citations.
Jupyter Notebook Expert Skill - Guide for notebook execution and Databricks kernel integration Use when: - Executing Jupyter notebooks via CLI (jupyter execute) - Running notebooks with Databricks kernel - Configuring notebook execution options (timeout, inplace)
In-depth Amazon keyword research and intelligent classification analysis. Collects 2,000+ keywords based on Sorftime MCP data, then uses an LLM Agent to classify them across 8 dimensions (negative words, brand words, material words, scenario words, attribute words, function words, core words, other), generating a Markdown report, a CSV keyword library, and an HTML dashboard. Trigger: /keyword-research {ASIN} {SITE}
This skill should be used when working with LaminDB, an open-source data framework for biology that makes data queryable, traceable, reproducible, and FAIR. Use when managing biological datasets (scRNA-seq, spatial, flow cytometry, etc.), tracking computational workflows, curating and validating data with biological ontologies, building data lakehouses, or ensuring data lineage and reproducibility in biological research. Covers data management, annotation, ontologies (genes, cell types, diseases, tissues), schema validation, integrations with workflow managers (Nextflow, Snakemake) and MLOps platforms (W&B, MLflow), and deployment strategies
Generate C/C++ or CUDA code from an AI model (PyTorch, LiteRT) using MATLAB Coder or GPU Coder. Use when the user wants to integrate an AI model into an application with code generation as the end goal — generating MEX, CUDA MEX, static library, dynamic library, or executable — or using the model in Simulink for simulation and code generation. Covers PyTorch ExportedProgram (.pt2) via loadPyTorchExportedProgram and LiteRT (.tflite) via loadLiteRTModel (R2026a+). Keywords: PyTorch, torch, .pt2, ExportedProgram, loadPyTorchExportedProgram, invoke, codegen, MEX, CUDA, GPU, C, C++, deploy, AI model, deep learning model, LiteRT, TFLite, TensorFlow Lite, Simulink, slbuild, PyTorch ExportedProgram block, MATLAB Function block, dlosslib, loadLiteRTModel.
Fit curves and surfaces interactively with the Curve Fitter app for a complete no-code fitting workflow. Invoke this skill when the Curve Fitter app, cftool, curveFitter, or "curve fitting tool/app" is mentioned in any way. Also use when exploring or comparing fit types (regression, interpolation, smoothing, splines, custom equations); excluding outliers interactively; iterating on a fitting workflow; help choosing a fit type; and exporting to a figure, generating MATLAB code, fits to the workspace, and to Simulink Lookup Tables.
Designs, reviews, and migrates MATLAB R2026a and GNU Octave numerical code, covering functions with arguments blocks, arrays and indexing, tables and timetables, matlab.unittest tests, MATLAB Projects, exportgraphics figures, MAT file inventory, and MATLAB-Python interoperability. Includes static helper scripts that scan .m files, plan argv commands, and hash artifacts without executing anything. Use when writing or reviewing .m code, migrating code between MATLAB releases or to Octave, setting up reproducible tests and projects, inspecting an untrusted MAT or .mlx file safely, or checking Python Engine compatibility. Do not use for pure Python numerical work with NumPy or SciPy.
Read, write, and manipulate medical imaging data (DICOM, NIfTI, NRRD) in MATLAB. Covers Image Processing Toolbox functions (dicomreadVolume, niftiread, dicomContours, dicomanon) and Medical Imaging Toolbox enhanced APIs (medicalVolume, medicalImage, medicalref3d, extractSlice, updateOrientation). Use when reading medical files, listing DICOM series, extracting spatial referencing, changing orientation, working with RT structures, or anonymizing DICOM data. Some features require Medical Imaging Toolbox — see skill body and references for details.
Battle-tested PyTorch training recipes for all domains — LLMs, vision, diffusion, medical imaging, protein/drug discovery, spatial omics, genomics. Covers training loops, optimizer selection (AdamW, Muon), LR scheduling, mixed precision, debugging, and systematic experimentation. Use when training or fine-tuning neural networks, debugging loss spikes or OOM, choosing architectures, or optimizing GPU throughput.
Identifies fantasy baseball players whose surface stats (wOBA, ERA, batting average) are diverging from their underlying Statcast quality (xwOBA, FIP, xBA) — emits a `regression_index` from -100 (very lucky, sell high) to +100 (very unlucky, buy low). Primary signal for buy-low/sell-high decisions on trades and waivers. Use when user mentions "buy low", "sell high", "regression candidate", "lucky", "unlucky", "xwOBA gap", "ERA-FIP gap", "BABIP", "due for regression", or is deciding whether to trade for / trade away a player based on over- or under-performance.
Create, revise, audit or polish submission-grade manuscript figures for Nature-family and other high-impact journals using Python or R. Covers multi-panel scientific plots, matplotlib/seaborn, ggplot2/patchwork/ComplexHeatmap and journal-ready SVG/PDF/TIFF exports. Also use for OpenRouter GPT Image 2 graphical abstracts, concept or mechanism schematics when the user explicitly requests AI generation. Establish the figure conclusion, evidence logic and exports before execution; honor the chosen or saved plotting backend. Trigger on paper figures, scientific/academic plotting, manuscript data visualization, 论文配图、学术写作配图、科研绘图、科研作图、画图、作图、出图、论文图表、可视化、论文示意图、机制示意图、图形摘要. Not for dashboards or Illustrator/Figma-first infographics.
Queries the NCATS Translator ARAX production API for bounded, typed, provenance-rich one-hop and endpoint-pinned two-hop biomedical knowledge-graph relationships. Use for Biolink-constrained RTX-KG2 lookup, explicit selected-provider ARAX federation, separate entity normalization, qualifier-aware graph traversal, and inspection of TRAPI edge bindings, publications, and knowledge-source provenance. Do not use for inference, ranking, open-ended pathfinding, clinical guidance, or sensitive queries
Guides and best practices for working with Neon Serverless Postgres. Covers getting started, local development with Neon, choosing a connection method, Neon features, authentication (@neondatabase/auth), PostgREST-style data API (@neondatabase/neon-js), Neon CLI, and Neon's Platform API/SDKs. Use for any Neon-related questions.
Provides guidance for interpreting and manipulating neural network internals using nnsight with optional NDIF remote execution. Use when needing to run interpretability experiments on massive models (70B+) without local GPU resources, or when working with any PyTorch architecture.
This skill should be used when user asks to "query OpenObserve", "create OpenObserve dashboard", "edit OpenObserve panel", "fetch OpenObserve logs", "run OpenObserve search", "list OpenObserve streams", "ingest into OpenObserve", or works with OpenObserve Cloud / self-hosted via REST API. Covers auth, search/SQL, streams, dashboards (CRUD + per-panel ops), the v8 panel JSON schema, and known pitfalls (re-aggregation, hash concurrency, microsecond timestamps).
Use when the user wants to iteratively improve an artifact under a hard correctness bound while minimizing a measured cost — refactoring a code module to cut complexity while its test suite stays green, OR speeding up a SQL query while it returns the same rows. Each iteration applies one focused change, checks a correctness gate that must pass, measures a metric that must drop, and keeps the change only if both hold, else reverts; loops to a plateau or budget. Not for adding features, fixing bugs, or any change that is allowed to alter behaviour or results.
Review and optimize application data access that uses Entity Framework Core (EF Core). Use for EF Core performance and correctness reviews of services, repositories, imports, and background jobs: slow or memory-heavy queries, excessive round-trips, hidden SaveChanges, premature materialization, dynamic query construction, result-graph tracking, bulk writes, DbContext lifetime or overlapping operations, search semantics, paging, raw SQL, and model/index coverage. Use whether or not EF Core owns the schema. For EF Core, not Dapper or raw ADO.NET.
Query PubMed publication trends via x402 micropayments to identify emerging research fields, compare therapeutic areas, detect publication spikes, find key papers, and profile authors. Use when analyzing biotech/pharma landscapes, writing grant applications, evaluating drug targets, or researching scientific trends. Requires the x402 skill from coinbase/agentic-wallet-skills for payment commands.
Bayesian statistical modeling with PyMC v5+. Use when building probabilistic models, specifying priors, running MCMC inference, diagnosing convergence, or comparing models. Covers PyMC, ArviZ, pymc-bart, pymc-extras, nutpie, and JAX/NumPyro backends. Triggers on tasks involving: Bayesian inference, posterior sampling, hierarchical/multilevel models, GLMs, time series, Gaussian processes, BART, mixture models, prior/posterior predictive checks, MCMC diagnostics, LOO-CV, WAIC, model comparison, or causal inference with do/observe.
Provides guidance for performing causal interventions on PyTorch models using pyvene's declarative intervention framework. Use when conducting causal tracing, activation patching, interchange intervention training, or testing causal hypotheses about model behavior.
Qdrant vector database: collections, points, payload filtering, indexing, quantization, snapshots, and Docker/Kubernetes deployment. Use when managing Qdrant collections, performing vector searches with payload filters, configuring HNSW indexes or quantization, or deploying Qdrant clusters. Keywords: Qdrant, vector database, HNSW, quantization, semantic search.
QMEDR EDR telemetry threat hunting: after the runtime metadata explicitly exposes hunt, view the Doris/OCSF schema and run read-only SQL with a UTC time window and LIMIT, supporting searches by EID, event UID, process command line, file or network clues. Does not cover investigations of Endpoint/Signal/Incident operational objects, policy management, general database administration or general SQL development.
Run a bounded red-teaming autoresearch loop to generate LLM guardrail training data. You (the agent running the skill) are the attacker and the judge: you craft attacks and label every response. The only model the harness calls is the target under test, over any OpenAI-compatible API (OpenRouter, Moonshot/Kimi, Fireworks, Ubicloud, OpenAI, or custom), with its key in `.red-team/.env`. Every attempt (pass and fail) is written to JSONL ready for fine-tuning guardrails in Llama Guard format. Use when asked to red-team or stress-test an LLM for harmful content, jailbreaks, prompt injection, or backdoor/trigger behavior; to build a safe/unsafe dataset; or to mine novel attacks at scale. Complements the hacker skill (code/infra exploitability): this skill targets model behavior and produces data.
[Deprecated] Use when building frontend Web apps that talk to CloudBase Relational Database via @cloudbase/js-sdk – provides the canonical init pattern so you can then use Supabase-style queries from the browser. New environments should use PostgreSQL with app.rdb() — see postgresql-development skill instead.
Parse and report retirement holdings from Vanguard and Fidelity CSV exports. CSV-only by design — the Vanguard IRAs and Fidelity 401k are not connected in SnapTrade, so there is no live/DB path yet and no persistent destination. Triggers on sync retirement, update retirement, vanguard sync, 401k update, IRA sync, or working with imports/retirement/ files.
Run Nextflow Registry modules natively using `nextflow module` commands. Use when running, listing, or getting info about Nextflow modules, including process modules (a single tool) and workflow modules (subworkflows that chain several modules).
DEPRECATED - Use browsing-bluesky skill instead. Sample and analyze Bluesky firehose to identify trending topics and content clusters. Use when user asks about "what's happening on Bluesky", "Bluesky trends", "zeitgeist", "firehose analysis", or wants to see real-time topic clusters from the network.
Builds, evaluates, and audits right-censored survival analysis workflows with scikit-survival (sksurv): Cox PH, Coxnet, IPC ridge, survival trees, forests, boosting, and SVMs, plus nonparametric cumulative incidence for competing risks. Covers leakage-safe scikit-learn pipelines, nested CV, and censoring-aware metrics. Use when fitting survival models on time-to-event data, building structured outcome arrays, computing IPCW concordance, dynamic AUC, or Brier scores, estimating cause-specific cumulative incidence, or tuning models without leakage. Not for Fine-Gray regression, which scikit-survival does not provide.
Video processing library for scientists. Provides easy access to video files using FFmpeg, motion estimation algorithms, and video quality metrics. Built on NumPy and designed for high-performance research in computer vision and image sequence analysis. Use when working with video files, motion estimation, video quality assessment (VQA), FFmpeg, temporal image data, video codecs, YUV data, or scientific video recordings.
Language-independent tokenizer treating text as raw Unicode. Supports BPE and Unigram algorithms. Fast (50k sentences/sec), lightweight (6MB memory), deterministic vocabulary. Used by T5, ALBERT, XLNet, mBART. Train on raw text without pre-tokenization. Use when you need multilingual support, CJK languages, or reproducible tokenization.
Use when someone wants to find product-data problems across a Shopify catalog: "audit my catalog", "find PDP problems", "which products have missing photos / thin or missing descriptions / pricing anomalies", "$0 or compare-at-inverted prices", "products missing type or vendor", "catalog quality sweep", "product data health check", "which listings are broken", "run a catalog audit before a big sale". Read-only: it locates and ranks problems, it does not change anything. Not for FIXING the problems it finds (use shopify-catalog-cleanup for debris/archiving, shopify-seo-metadata for missing SEO meta, shopify-alt-text for missing image alt text).
Provides guidance for training and analyzing Sparse Autoencoders (SAEs) using SAELens to decompose neural network activations into interpretable features. Use when discovering interpretable features, analyzing superposition, or studying monosemantic representations in language models.
Write correct, performant SQL across all major data warehouse dialects (Snowflake, BigQuery, Databricks, PostgreSQL, etc.). Use when writing queries, optimizing slow SQL, translating between dialects, or building complex analytical queries with CTEs, window functions, or aggregations
SQLAlchemy Python ORM. Covers models, sessions, and queries. Use for Python database applications. USE WHEN: user mentions "sqlalchemy", "declarative_base", "DeclarativeBase", "Session", "Query", "relationship()", "alembic", asks about "python orm", "sqlalchemy models", "sqlalchemy async", "sqlalchemy relationships", "python database" DO NOT USE FOR: Node.js/TypeScript ORMs - use `prisma`, `drizzle`, or `typeorm` skills; Django ORM - use Django-specific resources; raw SQL - use `database-query` MCP; NoSQL - use `mongodb` skill; other Python ORMs (Peewee, Pony) - not supported
Comprehensive guide for building solutions with the sqlx Rust library. Use this skill whenever the user wants to: write SQL queries in Rust with compile-time checking; connect to PostgreSQL, MySQL, or SQLite databases using async Rust; set up connection pools with sqlx; run database migrations; use query macros (query!, query_as!, query_scalar!); map SQL rows to Rust structs with FromRow; handle transactions in Rust; use #[sqlx::test] for database testing; work with JSON columns in SQL; use PgListener for PostgreSQL LISTEN/NOTIFY; handle sqlx errors properly; configure sqlx feature flags in Cargo.toml; set up offline mode with cargo sqlx prepare; use QueryBuilder for dynamic queries; or implement the repository pattern with sqlx. Make sure to use this skill when the user mentions sqlx, Rust database, Rust PostgreSQL, Rust MySQL, Rust SQLite, Rust SQL toolkit, compile-time query checking, async database Rust, database migrations Rust, sqlx pool, sqlx macros, or any SQL database interaction in Rust.
New to MCP?
From what the protocol is to installing your first server, these answers cover most first questions.
An open protocol that defines how AI applications connect to external tools and data in a uniform way. Clients talk to the model; MCP servers provide tools, resources and prompt templates. They exchange standard messages, so one server can be reused by many clients.