Skills
97,222 packaged capabilities and playbooks for agents. Install any of them with one command.
Data & databases
Use this skill for persistent PostgreSQL storage in agent workflows — spin up databases on demand, test schema migrations safely on forks, and pass databases as assets between agents using Ghost. **Trigger when user asks to:** - Create a new PostgreSQL database quickly - Manage database lifecycle in agent workflows - Fork a database to test changes safely - Run analytics or heavy queries on a fork without impacting the production database - Share a database snapshot so others can spin up their own copy - Pass a database as an input or output between agents - Create a database for agents - Need many databases that are billed only when queried - Set up the Ghost MCP server or CLI **Note:** If the Ghost MCP server is already installed, use its tools directly — this skill is for environments without the MCP or as a reference for what's possible. **Keywords:** Ghost, ghost.build, database, create database, fork database, PostgreSQL, managed Postgres, MCP, agent database, connection string
Invoke when Earth Engine, GEE, ee., or geemap is named; when work needs its server-side catalog; or when choosing Earth Engine versus local xarray or desktop processing for a large area or long archive. Covers image collections, masking, compositing, reducers, zonal statistics, time series, classification, quota-aware batching, and exports. This is an execution platform skill; combine it with remote-sensing-analysis or change-detection when those skills own the scientific method.
Trace relationships between semantic models and downstream reports across Fabric workspaces. Automatically invoke when the user asks to "find downstream reports", "show report lineage", "impact analysis", "what depends on this dataset", "cross-workspace lineage", "which reports are connected", "get model dependencies", or mentions model-to-report dependency tracing.
Use when creating, editing, or reviewing Liquibase database migrations in FOLIO Spring Boot modules. Always use this skill for any db migration, schema change, changeset authoring, XML changelog creation, database migration script, or table/column/index/FK changes in a FOLIO module — even if the user just says "add a column", "create a table", or "write a migration". Enforces FOLIO folder structure (changelog-master → changelog-vX.Y → vX.Y/ files), changeset ID naming convention (JIRA@@scope-action-target), comment tags, and FOLIO-standard column types.
Build 3D geological models with implicit surfaces, faults, folds, and stratigraphic constraints from structural geology data. Use when the agent needs to: (1) Build 3D geological models from structural data, (2) Model fault networks and displacements, (3) Create folded geology representations, (4) Interpolate geological surfaces, (5) Export models to VTK for visualization, (6) Perform uncertainty analysis on geological models, (7) Evaluate model values on grids.
Product validation skill of the lyt series for existing products. Use when the user already has a specific product, a candidate product list, product screenshots, 1688/competitor links, supply chain materials, a CSV/XLSX candidate product table or backend test data, and wants to judge whether a product or a batch of candidate products is worth entering the test pool. Outputs "can test / hold / not recommended", facts/inferences/assumptions, 7-dimension validation, a minimal validation plan and a data gap checklist. Only handles product validation questions related to cross-border e-commerce, TikTok Shop, e-commerce product sales, supply chain, product pools, content, advertising, influencer distribution and store operations. Triggers: /lyt-product-validation, /产品验证, /单品验证, /候选品筛选, "can this product be done" "how to filter these candidate products"
This skill should be used when the user asks to "calculate TAM", "determine SAM", "estimate SOM", "size the market", "calculate market opportunity", "what's the total addressable market", or requests market sizing analysis for a startup or business opportunity.
When the user wants to solve complex optimization problems using metaheuristics, apply genetic algorithms, simulated annealing, or other nature-inspired algorithms. Also use when the user mentions "genetic algorithm," "simulated annealing," "tabu search," "ant colony," "particle swarm," "evolutionary algorithms," "nature-inspired optimization," "heuristic search," or when exact optimization methods are too slow for large-scale problems. For exact methods, see optimization-modeling. For multi-objective, see multi-objective-optimization.
Create and edit Obsidian Bases (.base files) with views, filters, formulas, and summaries. Use when working with .base files, creating database-like views of notes, or when the user mentions Bases, table views, card views, filters, or formulas in Obsidian.
OpenBio API for biological data access and computational biology tools. Use when: (1) Querying biological databases (PDB, UniProt, ChEMBL, etc.), (2) Searching scientific literature (PubMed, bioRxiv, arXiv), (3) Running structure prediction (Boltz, Chai, ProteinMPNN), (4) Performing pathway/enrichment analysis, (5) Designing molecular biology experiments (primers, cloning), (6) Analyzing variants and clinical data, (7) Analyzing and editing plasmid files (GenBank, SnapGene).
Produce a read-only, evidence-based paid-media performance review across connected platforms or supplied exports. Use for weekly or monthly reports, scorecards, performance, CPA, ROAS, CTR, spend trends, pacing, or conversion-tracking health.
Performs pandas DataFrame operations for data analysis, manipulation, and transformation. Use when working with pandas DataFrames, data cleaning, aggregation, merging, or time series analysis. Invoke for data manipulation tasks such as joining DataFrames on multiple keys, pivoting tables, resampling time series, handling NaN values with interpolation or forward-fill, groupby aggregations, type conversion, or performance optimization of large datasets.
Build grounded question answering and retrieval-augmented generation (RAG) over your own collection of research papers, with answers that cite the exact paper and passage. Use when a user wants to "chat with" or search a folder of PDFs, synthesize evidence across a literature corpus, find which paper says X, or build a vector/hybrid index with SQLite FTS5, pgvector (Postgres), Chroma, Qdrant or FAISS. Includes a zero-dependency local full-text index with citable hits.
Interpret the meaning of a paper's figures and output a highly readable Markdown report that "teaches humans how to read the figures"; supports input of one or more absolute paths to figure files plus human interpretations, automatically attempts to retrieve the source code that generated each figure from near the figure, and interprets each figure with process-level isolation via `codex exec`/`claude -p` in a parallel-vibe-like manner (default concurrency limit 3, adjustable in config.yaml). Not applicable: the user merely wants to resize/crop/convert a figure, or asks to directly modify images or source code (this skill is read-only for images and source code throughout, and modification is strictly prohibited).
Keyword Tool data from eRank, plus local scoring, drift, and listing-gap analysis for Etsy sellers. Trigger phrases: `research dad mug on eRank`, `score an Etsy keyword`, `find eRank top listings`, `compare Etsy listing tags`, `use eRank`, `run eRank`.
Prisma ORM 7 CLI reference for existing Prisma 7 applications or explicitly selected Prisma 7 workflows. Covers prisma init, generate, migrate, db, dev, studio, and mcp. For setup, connection repair, or Prisma 8, use prisma-orm-setup; not the Prisma Platform CLI.
This skill MUST be used for product conjoint analysis (商品聯合分析、屬性偏好分析) to study trade-offs among product attributes using ratings, stated choices, or observed choices. It SHOULD be used for attribute importance, willingness-to-pay (願付價格), or product configuration questions when multi-attribute preference data is available. It MUST NOT treat review mentions as observed choices or produce pricing and probability claims unsupported by the data and fitted model.
Compare intended product outcomes against observed results to close the launch-to-learning loop: collect post-launch evidence, distinguish expected from observed from uncertain from inferred claims, update assumptions, assess feature health, and choose among continue/improve/harvest/pivot/pause/retire — including retirement lifecycles with deprecation, migration, customer treatment, and retained reusable learning. Do not use for incident postmortems or root-cause analysis (routes to incident-learning or site-reliability-engineering); do not use for analytics instrumentation or metric dashboard design (routes to product-analytics-and-measurement); do not use arbitrary thresholds as universal retirement rules — decisions require human judgment and context.
Multi-method geophysical modelling and inversion framework. Use when the agent needs to: (1) Perform electrical resistivity tomography (ERT) inversion, (2) Run seismic refraction tomography (SRT), (3) Model induced polarization (IP) data, (4) Simulate ground penetrating radar (GPR), (5) Create finite element meshes for geophysical problems, (6) Perform joint inversions of multiple datasets, (7) Forward model geophysical responses, (8) Analyze time-lapse monitoring data.
Use when a change spans more than one statement or file — a release or migration script, “chay script DDL”, “deploy ban release”, “run these .sql files”, a table with its comments and indexes, a package spec and body together, a directory of numbered scripts. Orders the statements, previews every one, gets one approval for the set, runs them through apply's six steps in order and stops at the first failure.
Use when designing or writing a new package, procedure, function or trigger, or restructuring one — “viet package”, “tao procedure”, “them function”, “them trigger”, an API another team will call, an error-handling or logging design, a spec change. Applies the industry-standard PL/SQL design rules — the spec as a contract, named exceptions, one transaction owner, thin triggers — beneath the house conventions, which always win.
Use when creating or changing a table, column, constraint, index, sequence or comment — “add a column”, “create a table for X”, “them cot”, “tao bang”, “sua kieu du lieu”, “comment cho cot”, a migration script, or a column type change. Designs the DDL to the house standard and Oracle practice, documents every table and column with COMMENT ON, and lands it through pythia deploy with the honest rollback.
Use when a query, report or PL/SQL unit is slow, before adding an index or a hint, or when someone says “cham”, “toi uu truy van”, “optimize this query”, “why is this report slow”, or “full table scan”. Measures with the optimizer's own plan and statistics, changes one thing at a time, and proves the gain with the same measurement — license-safe, no AWR/ASH or Tuning Pack.
Use before saying an Oracle task is done, fixed, deployed, faster or working — before the final report, before “xong roi”, “da sua xong”, “deployed”, before ending an autopilot session. Collects the evidence from this session — compile errors, INVALID objects against the baseline, lint, journal entries, before/after plans when performance was the point — and reports only what that evidence proves.
Guides Qdrant deployment selection. Use when someone asks “how to deploy Qdrant,” “Docker vs Cloud,” “local mode,” “embedded Qdrant,” “Qdrant EDGE,” “which deployment option,” “self-hosted vs cloud,” or “need lowest latency deployment.” Also use when choosing between deployment types for a new project
Guides embedding model migration in Qdrant without downtime. Use when someone asks “how to switch embedding models,” “how to migrate vectors,” “how to update to a new model,” “zero-downtime model change,” “how to re-embed my data,” or “can I use two models at once.” Also use when upgrading model dimensions, switching providers, or A/B testing models.
Guides Qdrant scaling decisions. Use when someone asks 'how many nodes do I need', 'data doesn't fit on one node', 'need more throughput', 'cluster is slow', 'too many tenants', 'vertical or horizontal', 'how to shard', or 'need to add capacity'.
Diagnoses and improves Qdrant search relevance. Use when someone reports 'search results are bad', 'wrong results', 'low precision', 'low recall', 'irrelevant matches', 'missing expected results', or asks 'how to improve search quality?', 'which embedding model?', 'should I use hybrid search?', 'should I use reranking?'. Also use when search quality degrades after quantization, model change, or data growth.
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).
Plan a web scrape and author a validated extraction spec — explore a detail page, discover fields, confirm the schema, then validate it with more pages, HTML-variant comparison, extracted values, and an optional browser review, and present the plan for approval.
Create web scraper scripts using Playwright. Use when the user wants to scrape a website, extract data from web pages, build a scraping pipeline, initialize a scraping project, or automate data collection from websites. Supports Python and Node.js ecosystems. Handles proxy configuration, stealth mode with patchright, and data export to CSV, JSON, or database.
Simulation and Parameter Estimation in Geophysics. Framework for geophysical forward modeling and inversion. Use when the agent needs to: (1) Run geophysical inversions (DC resistivity, magnetics, gravity, EM), (2) Create forward models for potential fields or electromagnetic methods, (3) Build survey geometries and receiver configurations, (4) Design mesh discretizations for simulations, (5) Apply regularization and optimization to inverse problems, (6) Model subsurface physical properties from geophysical data.
Master SQL query optimization, indexing strategies, and EXPLAIN analysis to dramatically improve database performance and eliminate slow queries. Use when debugging slow queries, designing database schemas, or optimizing application performance.
Postgres best practices maintained by Supabase, for Postgres running anywhere. Load this skill BEFORE writing or changing anything that lives in a Postgres database: creating or altering tables and columns (including choosing column types), schema design, migrations and declarative schema files, RLS policies and the tests that verify them, indexes, triggers, database functions, queues and scheduled jobs (pg_cron, pgmq), vector/semantic search (pgvector), and restoring dumps (pg_restore) or importing data. Also load it when diagnosing slow queries, high CPU, timeouts, EXPLAIN plans, connection exhaustion, locking, bloat, or rows visible to the wrong user or tenant. This is not just a performance guide — schema, migration, security, and SQL authoring tasks need these rules too, even for a one-column change or a single query.
Guide for implementing Syncfusion Windows Forms Pivot Grid control for data analysis and pivot table functionality. Use this skill when implementing pivot tables, data summarization, cross-tabulated data, or analytical dashboards in Windows Forms applications. Covers data binding, pivot configuration, filtering, sorting, grouping, calculations, conditional formatting, and exporting.
Calculate training costs for Tinker fine-tuning jobs. Use when estimating costs for Tinker LLM training, counting tokens in datasets, or comparing Tinker model training prices. Tokenizes datasets using the correct model tokenizer and provides accurate cost estimates.
Track AI token consumption, costs, and usage trends using the orbit CLI. Use this skill whenever the user asks about token usage, AI costs, Claude Code spending, how many tokens were used, cost breakdown by model, session history, or token analytics. Trigger on phrases like 'how much have I spent', 'token usage', 'show me costs', 'what's my AI spending', 'how many tokens today', 'cost per model', 'list sessions', 'track usage', 'token report', 'weekly usage', 'monthly costs', or any token/cost tracking task — even casual references like 'am I spending too much on Claude', 'what did that session cost', 'show me the dashboard', or 'how much is opus costing us'.
Interpret YouTube Analytics, TikTok Analytics, and video performance data. Identifies trends, explains metrics, and provides actionable recommendations for growth. Use when analyzing video performance, understanding metrics, or optimizing channel strategy.
Video Channel content search tool. Search popular Video Channel posts using keywords provided by the user, with sorting by newest, most likes, most saves, or relevance. Results are displayed in a structured table. Use when users want to find popular Video Channel content, search for viral videos, query post data, or learn how a type of content performs on Video Channel. Trigger phrases include: Video Channel search, Video Channel post lookup, viral Video Channel posts, popular Video Channel posts, Video Channel content search, search Video Channel, WeChat Video Channel search. When the API returns insufficient points (code=3210), automatically launch Alipay to top up and continue the search.
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.