Skills
97,501 packaged capabilities and playbooks for agents. Install any of them with one command.
Data & databases
Adds and runs data quality checks (dbt-test style assertions) on a project's warehouse tables and saved-query views, and HogQL catalog metrics: not-null, uniqueness, accepted values, referential integrity, row-count bounds, freshness, and custom HogQL. Metrics support custom SQL checks only. Use when asked to test a model, validate a view, check for nulls or duplicates, add data quality checks, find out why a number looks wrong, or judge whether a warehouse table is trustworthy before using it in an analysis. To describe what data *means* (metrics, certifications, joins), see setting-up-data-catalog instead. Trigger terms: data quality, data test, dbt test, not null check, uniqueness check, freshness check, referential integrity, row count check, validate model, is this table trustworthy.
Creates product analytics or SQL-backed box plot insights in PostHog. Use when a user asks to create, build, or save a box plot, visualize a numeric distribution, compare quartiles or medians across dates or groups, or turn SQL results into a box plot. Chooses between a standard Trends box plot and a SQL insight, validates the distribution data, saves the insight, and verifies it.
Looker Developer Onboarding: Step 6. Creates LookML views and models based on the user's goals, maps model connections, validates LookML, and runs verification queries. Only execute this after Step 5 (Project Setup using `setting-up-looker-project`).
Reference for the verbose watsonx.data integration SDK for DataStage (batch) flows, with exhaustive stage and property access — the full engine stage catalog. Load this only after the di-agent-flow-lifecycle router has selected the datastage-sdk authoring backend — this is a syntax and stage reference, not a decision about whether to use it. If you have not routed yet, load di-agent-flow-lifecycle instead.
API spec for pyflow, IBM's LLM-optimized Python DSL for authoring DataStage and StreamSets flows. Its compact surface and compile-time validation offer context efficiency, fast feedback, and correctness guarantees. Load this only after the di-agent-flow-lifecycle router has selected the pyflow authoring backend — this is a language reference, not a decision about whether to use it. If you have not routed yet, load di-agent-flow-lifecycle instead.
Enables Customer-Managed Encryption Keys (CMEK) on CockroachDB Cloud clusters with the Advanced plan and Advanced Security Add-on to give organizations control over data-at-rest encryption keys via their cloud provider's KMS. Use when enabling CMEK for compliance, rotating encryption keys, or verifying CMEK configuration.
Looker Developer Onboarding: Step 1. Guides the agent to explore BigQuery data and define the onboarding goal. This is the first active step, to be executed immediately after reading the parent orchestrator `looker-developer-onboarding`.
Build a starting-point taxonomy for an MCP tool — what users were trying to accomplish before they reached the tool — and publish it as a PostHog notebook. Reconstructs each session's goal from its opening tool calls, then clusters those goals into named categories with size, share, and facet mix. Use when the user asks "why do people use this tool?", "what are users actually trying to do?", "what problem brings people here?", "where do these sessions start?", "segment usage of <tool> by goal", or wants a Clio-style taxonomy of MCP usage. Complements exploring-mcp-intent-clusters, which groups what agents did per call rather than why the session began. The agent running this skill writes the goal labels itself, reading the corpus query output session by session — the bundled scripts cover the mechanical facets but measurably lose the goal's altitude, so do not delegate that field to them.
Use when implementing Redis caching, cache invalidation, or distributed locking in Frappe. Prevents stale cache bugs, race conditions from missing locks, and memory bloat from unbounded cache keys. Covers frappe.cache(), @redis_cache decorator, cache.get_value/set_value, cache invalidation patterns, frappe.lock, TTL strategies. Keywords: cache, Redis, redis_cache, invalidation, locking, frappe.cache, get_value, set_value, TTL, distributed lock, data not refreshing, stale data, cache not clearing, Redis error, slow repeated queries..
Comprehensive SQL statement checking for ClickHouse, supporting multiple kernel versions (24.8, 23.3, 22.3) and two check modes: 1. Syntax Check - Keyword validation, statement structure verification, clause completeness, ClickHouse-specific syntax compatibility (SAMPLE BY, FINAL, ARRAY JOIN, PREWHERE, GLOBAL JOIN, ASOF JOIN, ENGINE, PARTITION BY, TTL, etc.) based on kernel source grammar 2. Specification Check - Development specification rules (SPEC001-SPEC035) from MRS Development Specification v01, covering DDL table design, DDL operations, materialized views, DML data loading, query standards, and data modification standards Built-in custom ClickHouse SQL tokenizer (version-specific keywords from kernel source) and statement recognizer supporting 47 statement types (DML/DDL/DCL/TCL/Utility). Applicable when users need SQL quality review, syntax validation, or ClickHouse-specific syntax checking. Trigger: "Clickhouse SQL check", "CK SQL check", "Clickhouse SQL 校验", "Clickhouse SQL 检查", "Clickhouse SQL specification", "Clickhouse SQL 规范检查", "Clickhouse SQL optimization", "Clickhouse SQL 优化", "检查 Clickhouse SQL"
Best practices for using vectors and AI with MariaDB. Use when writing SQL involving VECTOR columns, building RAG or semantic search with MariaDB, choosing embedding models, designing vector indexes, or integrating MariaDB with AI frameworks like LangChain, LlamaIndex, or Spring AI. Also use when the user mentions MariaDB and any of: vectors, embeddings, similarity search, nearest neighbor, AI, RAG, or LLM. IMPORTANT — MariaDB has native built-in vector support since version 11.7; it is NOT a plugin or extension (unlike pgvector for PostgreSQL). Do not recommend installing any extension.
Using the Materialize Terraform provider to manage Materialize resources declaratively. Covers clusters, sources (Kafka, Postgres, MySQL, SQL Server), sinks (Kafka, Iceberg), connections, materialized views, indexes, tables, roles, grants, secrets, network policies, and cloud-only resources (users, SSO, SCIM, app passwords). Use this skill whenever the user asks about writing Terraform for Materialize, creating or configuring Materialize resources with Terraform, importing existing Materialize objects into Terraform state, configuring the Materialize provider for Cloud or self-managed, setting up RBAC or grants via Terraform, creating connections or sources in Terraform, or troubleshooting Terraform plan/apply issues with Materialize resources. Also trigger when the user mentions materialize_cluster, materialize_source_kafka, materialize_connection_postgres, or any other materialize_* resource type.
Guide for implementing Grafana Mimir - a horizontally scalable, highly available, multi-tenant TSDB for long-term storage of Prometheus metrics. Use when configuring Mimir on Kubernetes, setting up Azure/S3/GCS storage backends, troubleshooting authentication issues, or optimizing performance.
Design, implement, and validate reproducible machine-learning pipelines spanning data preparation, training, evaluation, registry, and deployment gates. Use when the user requests an ML pipeline, needs to turn model scripts into an orchestrated workflow, or provides pipeline components that must be connected safely.
Guide for using the CockroachDB replicator to continuously replicate changes from PostgreSQL, MySQL, or Oracle to CockroachDB after an initial molt fetch data load. Use when setting up CDC replication, configuring pglogical/mylogical/oraclelogminer, or managing the fetch → replicator cutover workflow.
MyBatis-Plus enhanced ORM skill. Covers choosing between LambdaQueryWrapper and QueryWrapper (always Lambda), pagination plugin (PaginationInnerInterceptor) configuration, optimistic locking (@Version + OptimisticLockerInnerInterceptor), logical deletion (@TableLogic global/local configuration), auto-fill (@TableField fill + MetaObjectHandler), choosing between ActiveRecord and Mapper modes, preventing full-table updates (BlockAttackInnerInterceptor), and extracting a generic PageQuery. Use when the user performs database operations with MyBatis-Plus, configures pagination or optimistic locking, or chooses a query approach. Complements the mybatis skill: mybatis focuses on XML/ResultMap/SQL style, while mybatis-plus focuses on enhanced features.
Get web data now — fast, incremental, immediately responsive to what the user needs. The only way Claude can access live websites. USE FOR: - Fetching any URL or reading any webpage - Scraping prices, listings, reviews, jobs, stats, docs from any site - Running Extraction Templates — reusable, site-specific structured scrapers - Running Web Search Agents — open-ended research, enrichment, and dataset building with citations - Discovering URLs on a site before bulk extraction - Calling public REST/XHR API endpoints - Web search and research (8 focus modes) - Bulk crawling website sections Must be pre-installed and authenticated. Run `nimble --version` to verify (>= 1.2.0).
Get PostHog data into a canvas correctly: the host-injected `ph` SDK (loadInsight, query, capture, state, connectors, openExternal, navigate), the data hierarchy (saved insights first, typed query nodes second, inline HogQL last), verifiability (insight-backed metrics link their saved insight in PostHog; ad-hoc queries expose the exact query that ran), per-insight-type result shapes, progressive per-query loading, date-range wiring, live third-party data through the viewer's own connections (ph.connectors), and event capture from a canvas. Use whenever a canvas shows metrics, charts, tables, any PostHog data, or data from GitHub or an MCP server, or needs to send analytics events.
Provisions a Replay Vision `experiment` scanner that watches one experiment’s exposed sessions, records each session’s variant, samples variants evenly, and stops when the experiment ends. Covers status guards, a prompt that remains comparable across variants, sizing credit spend against the experiment’s own population, creating it disabled so its prompt can be previewed on real sessions, and reading results by variant. TRIGGER when: the user wants Replay Vision to watch an experiment, asks to scan or analyze an experiment’s recordings with AI, asks “what are users actually doing in the test variant,” or wants a scanner scoped to an experiment’s exposed sessions. DO NOT TRIGGER when: creating a general-purpose scanner not tied to an experiment (use creating-replay-vision-scanners), reading observations a scanner already produced (use exploring-replay-vision-observations), or manually browsing an experiment’s recordings without AI analysis (use analyzing-experiment-session-replays).
Populate person or group properties from a data warehouse table or materialized view, so warehouse columns become properties usable in feature flags, cohorts, and insights. Use when the user wants to "sync my Postgres columns to person properties", "map a warehouse table to people", "replace our daily identify cron", "update person properties from the warehouse", or asks for reverse ETL into PostHog. Covers picking the table and the identifier column, proposing a whole-table column mapping in one call, excluding identity and reserved properties, checking collisions with properties that already exist, and running the first backfill.
After data collection is complete, run confirmatory statistical analysis according to the analysis plan locked in before collection. Use when the user wants to compare groups (t-test/ANOVA/nonparametric tests), compare paired before-and-after measurements, run chi-square on contingency tables, do correlation analysis, calculate effect sizes and confidence intervals, apply multiple-comparison correction, check test assumptions (normality/homogeneity of variance), or write statistical results into reportable conclusions. Covers only "post-collection confirmatory analysis"; pre-collection design and sample size belong to the experimental design skill, and machine learning modeling belongs to the data processing skill.
Runs and reasons about the automated AI health check that suggests path-cleaning rules for web-analytics teams. Use when asked to generate path-cleaning suggestions for a team or cohort, to run the suggestion check, to review/apply AI-suggested rules, to inspect path_cleaning_suggestions health issues, or to extend the suggestion pipeline. Covers the suggest_path_cleaning_rules management command, the path_cleaning_suggestions health check, the cohort gating (precompute teams), and how suggestions are validated against real paths before storage. For hand-authoring or applying rules directly, use managing-path-cleaning-rules instead.
Post-market safety surveillance and recall/adverse-event RETRIEVAL across the full spectrum of FDA-regulated products that are NOT covered by the drug-AE signal skills: medical devices, food / dietary supplements / cosmetics, veterinary drugs, and drug supply (shortages). Orchestrates openFDA endpoints (MAUDE device adverse events + device recalls + 510(k), CAERS food/supplement/ cosmetic adverse events, veterinary adverse events, drug shortages, and cross-product enforcement/recall reports). USE WHEN the user asks: "are there adverse events for [device / pacemaker / infusion pump / insulin pump]", "device recalls for [firm/product]", "supplement / vitamin / cosmetic adverse reactions", "is [drug] in shortage", "what injectables are on shortage", "veterinary / animal adverse events for [drug] in [dog/cat/horse]", "food recall for listeria", "MAUDE report for [device]", "CAERS reactions for [brand]". DO NOT USE for drug adverse-event SIGNAL detection or disproportionality (PRR / ROR / IC) or drug-AE association scoring — that is `tooluniverse-pharmacovigilance` / `tooluniverse-adverse-event-detection`. This skill is multi-product surveillance and retrieval, not drug-AE statistical signal mining.
Explains PostHog billing usage and spend from the customer's visible Billing MCP tools. Use when the user asks why usage or spend is high, which product or project is driving usage, what a usage type means, how to reduce usage, what changed over time, why they got a usage change alert, or whether a spike/drop alert was real or noisy. Also use before product-specific analytics skills when the user names a billable PostHog product metric such as events, recordings, feature flag requests, exceptions, survey responses, synced rows, logs, AI events, AI credits, or Inbox credits. Starts from Billing usage/spend tools, then routes to customer-visible product MCP surfaces for deeper investigation.
Fine-tune LLMs with Unsloth using GRPO or SFT. Supports FP8, vision models, mobile deployment, Docker, packing, GGUF export. Use when: train with GRPO, fine-tune, reward functions, SFT training, FP8 training, vision fine-tuning, phone deployment, docker training, packing, export to GGUF.
Use when 王建硕 wants to systematically improve his X (Twitter) content by iterating on the content-generation prompt (prompts/x/prompt.md, used by the every-6h tweet Action) and finding which prompt version produces the highest-reach tweets. Each prompt edit is a git-SHA-versioned, numbered experiment with a hypothesis; tweets are attributed to the version live at post time and judged on median impressions per tweet. Also mines per-tweet impression data for content-feature signals (angle / length / topic) that feed the next prompt edit. North-star = impressions per tweet. Triggers — "改 X 的 prompt", "X 内容改进", "哪版 prompt 最好", "什么内容 impression 高", "improve my tweets", "A/B test the X prompt", "/wjs-x-improving-content".
Use when 王建硕 wants to systematically grow his X (Twitter) followers by running numbered, A/B-testable growth experiments and tracking which ones actually work. Every action gets a number, a hypothesis, a target metric, a before-state (for rollback), and a verdict. The North-Star metric is the new-follower ÷ profile-visit ratio (conversion) — every profile change is judged against it. Daily it ingests the X Analytics CSV export, scores each running experiment, and recommends keep / rollback. Triggers — "涨粉", "增加 X 粉丝", "X 涨粉实验", "follower growth", "A/B test my profile", "今天的涨粉检查", "/wjs-x-increasing-follower".
Work with PostHog Signals scouts: scheduled agents that monitor a project and write reports into the Signals inbox. Use to assign monitoring work, schedule quality scoring, find which scout covers a surface, act on reports, reduce noise, investigate missing findings, or improve the fleet through feedback and calibration. Also covers follow-up checks that verify whether a reported problem stays fixed. Use `authoring-scouts` for edits, `exploring-scouts` for run observability, and `inbox-exploration` for report triage. Trigger on "work with my scouts", "get more out of scouts", "have a scout watch X", "tell me if Y spikes", "score X on a schedule", "measure quality of Y", "what do I do with this scout report", "calibrate/review my scout fleet".
Use when running Alembic database migrations on top of a SQLAlchemy 2.x data layer that targets more than one dialect — SQLite, PostgreSQL, and MySQL — from a single migration history. Covers wiring env.py / alembic.ini to your models' Base.metadata (sourcing the DB URL from the environment, sync-first with a short async run_async_migrations note), the full revision -> autogenerate -> review -> upgrade/downgrade workflow, branching (history / current / heads / merge for multiple heads), and the load-bearing multi-dialect gotcha: SQLite cannot ALTER most schema, so write migrations in BATCH (move-and-copy) style — the same batch_alter_table code emits plain ALTER on PG/MySQL. Also owns the create_all-vs-Alembic decision and the baseline-stamp path off a create_all'd DB. Multi-dialect, portable, sync-first. Not SQLAlchemy model/engine authoring (see sqlalchemy), not job-queue/lease patterns (see sql-job-queue).
Use when creating Datex Studio datasources with dxs datasource/configuration commands: generating OData or flow datasource configs — both the cloud `-datasource.json` (typeId 6, called by functions/selectors) and the server-tier FootprintDatasource variant (`-footprintDatasource.json`, typeId 19, called by actions) — validating configs against a branch, and upserting standalone or owned datasources. Triggers: "create a datasource", "OData/flow datasource", "FPDS", "footprintDatasource", "a datasource an action can call", "standalone vs owned datasource".
When the user wants help with Google Ads attribution models, data-driven attribution, attribution windows, cross-channel attribution, how attribution affects Smart Bidding, multi-touch conversion paths, or understanding which campaigns are actually driving conversions. Triggers on 'attribution', 'attribution model', 'data-driven attribution', 'last click attribution', 'assisted conversions', 'conversion window', 'multi-touch', 'attribution model comparison', 'first click attribution', 'credit allocation', or 'which campaign is driving conversions'. For conversion tracking setup see google-ads-conversion-tracking. For Smart Bidding strategy see google-ads-bidding.
Agent-callable Google Ads tools — search campaigns, ad groups, and ads via GAQL, build performance reports, manage campaign status and budgets, and set up conversion tracking. Use when the user mentions Google Ads or wants to read or manage advertising campaigns, budgets, or reporting — even if they don't name Google Ads explicitly.
Using the mz-deploy CLI to manage a declarative SQL project for Materialize — project layout and the stage/wait/promote deployment lifecycle. Use this skill whenever the user is working in an mz-deploy project (a directory containing project.toml), asks how to deploy SQL changes to Materialize safely, or mentions mz-deploy, project.toml, profiles.toml, types.lock, staging deployments, deploy IDs, `mz-deploy stage`, `mz-deploy promote`, `mz-deploy abort`, `mz-deploy apply`, `SET api = stable`, stable API schemas, replacement materialized views, EXECUTE UNIT TEST, or per-profile SQL file overrides (`name#profile.sql`). Also trigger when a user needs to roll back a Materialize deployment or resolve a deployment conflict.
Use when a project's house style already exists somewhere and needs to become something the tooling can check - the developer hands over a standards document, points at an older or company base schema whose naming should be adopted, says "our team does it this way", or asks why an apply preview warned about a name. Derives the patterns from the real schema, verifies them against it, and writes .pythia/conventions.json and conventions.md.
Produces a standardized requirements brief from any source: DevOps work items, mockups, natural language descriptions, existing reports, or documents. This is a utility skill — it gathers and structures requirements but does not build anything. If the goal is to create a report, datasource, or other artifact, use the appropriate creation skill as the entry point; it will invoke this skill when it needs requirements.
Deep technical analysis skill for A-share holdings. Activates automatically when the user sends a screenshot of holdings or mentions intents related to holdings/stocks/investment advice. Activation conditions (any one triggers it): - The user sent an image (holdings screenshot, quote screenshot, etc.) - The user says "分析持仓", "我的持仓怎么样", "给投资建议", "持仓诊断", "我的股票", "目前持仓" - The user says "更新持仓", "录入持仓" - The user says "每日一股" or similar - Any technical analysis request involving holdings lookup, investment advice, or holdings diagnosis After activation: 1. If an image is received → prefer AI multimodal capability to read the screenshot content directly 2. If there is no image but there is holdings intent → read positions_portfolio.json 3. If there is no holdings data → inform the user and guide data entry 4. Run full technical analysis (AKShare real-time quotes + RSI/MACD/KDJ/Bollinger Bands/moving averages) 5. Generate action recommendations (hold/reduce/add/take profit/stop loss) 6. Output the analysis results directly in the conversation
A user-empowering data visualization Vue 3 components library (69 components) for eloquent data storytelling. ALWAYS use when writing code importing "vue-data-ui". Consult for component choice, config, theming, slots, tooltips, exports, SSR, debugging, or modifying vue-data-ui, vue data ui.
Use for semantic work on an existing Iskron graph: closing lifecycles, sense on arrows, splitting kriyas with two actors, re-hooking after distinguishing phenomena — the graph exists structurally but is semantically incomplete. Triggers: «проткать», «прошить», «ткачество», weave, close lifecycle, fix tensions, leaked/relay-gap/orphan, iskron_orient(lens=\"tensions\") shows problems, high tension score on entry. Distinct from design: design creates structure from goals, weaving completes an existing one.
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.