Skills
96,981 packaged capabilities and playbooks for agents. Install any of them with one command.
Data & databases
Crawl and export WeChat official account articles. Automatically handles login state acquisition and renewal for mp.weixin.qq.com, and supports searching by official account, crawling article lists and bodies, and exporting to Markdown / JSON / CSV by date window. Trigger when the user wants to crawl a WeChat public account, export recent articles, or mentions "wcx", "微信公众号", "公众号文章", "mp.weixin", "抓公众号", "crawl wechat official account", "wxmp", "最近十天的文章".
An A-share stock picking and trading plan skill based on "moving averages set the direction, MACD sets the rhythm". Use when the user asks to combine moving averages with MACD for stock picking, screening, pre-market candidate pools, trend following, pullback re-entry, golden cross confirmation, reducing positions on a bearish divergence, or wants to turn technical judgment into executable rules and risk-control templates.
Computes P(we win at least K of N categories) for a head-to-head categorical matchup via Monte-Carlo simulation or Poisson-binomial approximation. Domain-neutral — works for any fantasy sport with H2H Categories scoring (MLB, NBA, NHL) or any zero-sum per-category competition. Use when you need matchup_win_probability, per_cat_win_probability, expected_cats_won, or variance_estimate; or when user mentions "matchup win probability", "head to head simulation", "Monte Carlo matchup", "Poisson binomial matchup", "P win 6 of 10", "category matchup simulation", or "weekly win probability".
Use when the user wants to train, compare, or export machine learning models using Classification Learner or Regression Learner — including opening the app, loading data, training models, evaluating metrics, comparing results, visualizing plots, testing on held-out data, exploring model interpretability, and exporting trained models. Programmatic access to Classification Learner and Regression Learner apps via AppController.
mdtraj molecular dynamics trajectory analysis (Python). Reads DCD/XTC/TRR/NetCDF/H5/PDB topologies and trajectories; computes RMSD vs time, radius of gyration, per-residue RMSF, residue-residue contact frequency maps, phi/psi torsions for Ramachandran plots (general + Gly/Pro), and 8-state DSSP secondary structure. Modules: trajectory I/O, geometry (distances/angles/dihedrals), structural analysis (RMSD/Rg/RMSF/SASA), contacts, hydrogen bonds, secondary structure (DSSP), NMR observables. For broader atom-selection grammar use mdanalysis-trajectory; for running MD simulations use OpenMM/GROMACS.
Guided workflow for taking any ML codebase — including one with no experiment tracking at all, or one full of MLflow 2-era idioms — to a production-grade open-source MLflow 3 setup with dev/staging/prod environments, registry-based promotion, and served models. Walks seven phases with a developer who may have zero MLflow 3 experience — assess the codebase (scripted read-only audit), model the registry domain (per-environment model names, aliases, gates), stand up tracking per environment, restructure training code to MLflow 3 idioms, wire evaluation-gated promotion, serve and smoke-test, then run the ongoing MLOps loop. Use when asked to set up MLflow, migrate to MLflow 3, productionize model training and serving, or design a dev/staging/prod MLOps cycle. Pairs with the sibling mlflow-3 rule pack for every API decision.
Generate read-only MongoDB queries (find) or aggregation pipelines using natural language, with collection schema context and sample documents. Use this skill whenever the user asks to write, create, or generate MongoDB queries, wants to filter/query/aggregate data in MongoDB, asks "how do I query...", needs help with query syntax, or discusses finding/filtering/grouping MongoDB documents. Also use for translating SQL-like requests to MongoDB syntax. Does NOT handle Atlas Search ($search operator), vector/semantic search ($vectorSearch operator), fuzzy matching, autocomplete indexes, or relevance scoring - use search-and-ai for those. Does NOT analyze or optimize existing queries - use mongodb-query-optimizer for that. Does NOT handle aggregation pipelines that involve write operations. Requires MongoDB MCP server.
Help with MongoDB query optimization and indexing. Use only when the user asks for optimization or performance: "How do I optimize this query?", "How do I index this?", "Why is this query slow?", "Can you fix my slow queries?", "What are the slow queries on my cluster?", etc. Do not invoke for general MongoDB query writing unless user asks for performance or index help. Prefer indexing as optimization strategy. Use MongoDB MCP when available.
Set up serverless Postgres with Neon or Vercel Postgres for Cloudflare Workers/Edge. Includes connection pooling, git-like branching for preview environments, and Drizzle/Prisma integration. Use when: setting up edge Postgres, configuring database branching, or troubleshooting "TCP not supported", connection pool exhausted, SSL config (sslmode=require), or Prisma edge compatibility.
When the user wants to optimize network flows, solve minimum cost flow problems, or design flow networks. Also use when the user mentions "min cost flow," "maximum flow," "network flow problem," "transportation problem," "transshipment problem," "multi-commodity flow," "supply chain flow optimization," or "network capacity planning." For facility location, see facility-location-problem. For distribution networks, see distribution-center-network.
Comprehensive toolkit for creating, analyzing, and visualizing complex networks and graphs in Python. Use for network/graph data structures, analyzing relationships between entities, computing graph algorithms (shortest paths, centrality, clustering), detecting communities, generating synthetic networks, or visualizing network topologies. Applicable to social networks, biological networks, transportation systems, citation networks, and any domain involving pairwise relationships.
A skill that queries data by running SQL directly against the development database. Triggered by the keywords "DB 확인", "데이터 조회", "테이블 조회", "SQL 실행", "데이터 검증". Use when an AI agent needs to immediately check development database data during backend/frontend development.
Pipeline analysis composite. Pulls deal/meeting data from any CRM or tracking system, analyzes the pipeline over a user-defined period (weekly, fortnightly, monthly, quarterly), and produces both an executive summary and a detailed diagnostic report. Covers volume, qualification rates, source effectiveness, stage velocity, stuck deals, and actionable recommendations. Tool-agnostic — works with any CRM (Salesforce, HubSpot, Pipedrive, Close, Supabase, CSV).
Use when operating Ultralytics Platform from a terminal or script with the ul cloud CLI (ul cloud RESOURCE OPERATION key=value, PyPI package ultralytics-platform) — listing, creating, updating, cloning, or moving projects, datasets, and models; inspecting dataset images, statistics, metadata, duplicates, and similar images; starting and monitoring cloud training runs; per-image model validation analysis; downloading weights; exports, deployments, uploads, trash and restore, storage integrations, and account usage, storage, and billing; and building Agents (YOLO, LLM, gate, and alert workflows). For the Platform web UI or local yolo commands, see yolo; for choosing what to train and how to improve it, see yolo-training and yolo-tuning.
Design and review Polygres retrieval using available read-only MCP evidence across relational, graph, text, hybrid, Polygres AI Context, and existing pgvector strategies. Use for strategy selection, Context or graph design, grounded RAG planning, compatibility, and architecture review before implementation.
Diagnose Polygres MCP, OAuth connections, CLI, Runtime API, projects, jobs, synchronization, migrations, and retrieval through public read-only evidence. Use for scope or project-boundary issues, timeouts, partial failures, readiness, Context operations, graph, vector, text, hybrid, and Joint retrieval.
Query your local Chrome browsing history — search, topic clusters (Chrome's own Journeys), time/productivity reports, session timelines, downloads, and a behavioral profile — all read-only and on-device. Trigger phrases: `what have I been browsing`, `search my chrome history`, `what was that page I saw`, `my browsing journeys / topics`, `what did I download in chrome`, `what sites do I visit most`, `my browsing time report`, `my browsing profile`, `what was I researching last week`, `use chrome-history`, `run chrome-history-pp-cli`.
Use when building tabular machine learning models, making predictions on structured data, or integrating ML into applications. Reach for TabPFN when you need fast, accurate predictions without extensive preprocessing or hyperparameter tuning, especially for classification, regression, forecasting, or anomaly detection tasks.
Comprehensive healthcare AI toolkit for developing, testing, and deploying machine learning models with clinical data. This skill should be used when working with electronic health records (EHR), clinical prediction tasks (mortality, readmission, drug recommendation), medical coding systems (ICD, NDC, ATC), physiological signals (EEG, ECG), healthcare datasets (MIMIC-III/IV, eICU, OMOP), or implementing deep learning models for healthcare applications (RETAIN, SafeDrug, Transformer, GNN).
Guide for writing correct Bright Data Python SDK code. Always use this skill when writing, modifying, debugging, or reviewing Python code that uses the brightdata-sdk package, imports from brightdata, or interacts with Bright Data APIs. Use when the user asks to scrape websites, search Google/Bing, access datasets, or automate browsers via Bright Data in Python.
Analyze image files, image URLs, PDF files, and PDF URLs when the task requires recognizing, understanding, extracting, answering questions about, or locating content from the visual input. Typical uses include document parsing, layout analysis, element recognition, OCR, key information extraction, chart understanding, and document VQA. Do not use for plain text, structured data, code files, image-processing tasks, or cases where the needed information is already available in text form.
LLM workloads on open-source Ray (pinned to 2.57) — OpenAI-compatible serving with ray.serve.llm (vLLM-backed LLMConfig + build_openai_app) and batch inference with ray.data.llm (build_processor). Corrects the stale defaults a model produces — the archived ray-llm repo and its YAML configs, hand-rolled vLLM engines inside plain Serve deployments, the removed build_llm_processor name, deprecated boolean stage flags, top-level LLMServer/LLMRouter imports, free-form accelerator strings, one-deployment-per-LoRA-adapter designs — with the 2.57 idioms (stage configs, placement_group_config over hand-rolled PGs, deployment_config autoscaling, dynamic LoRA multiplexing, prefix-cache-affinity routing, the full OpenAI endpoint surface). Use when writing, reviewing, or productionizing LLM serving or batch inference on Ray. Classic-ML Ray (Train/Tune/Data/Serve/clusters) lives in the sibling ray skill.
Cheminformatics toolkit for fine-grained molecular control. SMILES/SDF parsing, descriptors (MW, LogP, TPSA), fingerprints, substructure search, 2D/3D generation, similarity, reactions. For standard workflows with simpler interface, use datamol (wrapper around RDKit). Use rdkit for advanced control, custom sanitization, specialized algorithms.
When the user wants to apply reinforcement learning to supply chain problems, learn optimal policies, or solve sequential decision-making under uncertainty. Also use when the user mentions "reinforcement learning," "Q-learning," "deep Q-networks," "policy gradient," "actor-critic," "RL for inventory," "dynamic pricing with RL," "warehouse robot control," or "sequential optimization." For forecasting, see neural-networks-forecasting. For static optimization, see optimization-modeling.
Design and audit release health dashboards, rollout gates, canary metrics, crash/error monitoring, support signals, store release status, desktop auto-update adoption, rollback triggers, and post-release review for apps, SaaS, games, web, and desktop software. Use when launching releases, patches, experiments, or staged rollouts that need operational confidence.
Use for Roblox persistent data and cross-server state design: choosing between DataStoreService, OrderedDataStore, MemoryStoreService, and MessagingService; designing save and load flows, schema shape, versioning, metadata, retries, quotas, observability, and concurrency-safe coordination across servers.
Simple Preference Optimization for LLM alignment. Reference-free alternative to DPO with better performance (+6.4 points on AlpacaEval 2.0). No reference model needed, more efficient than DPO. Use for preference alignment when want simpler, faster training than DPO/PPO.
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.