Skills
101,174 packaged capabilities and playbooks for agents. Install any of them with one command.
All
Create, analyze, and visualize complex networks and graphs in Python with NetworkX. Use when working with network/graph data structures, computing graph algorithms (shortest paths, centrality, clustering), detecting communities, generating synthetic networks (random, scale-free, small-world), reading/writing graph file formats, or drawing network topologies. Common applications include social, biological, transportation, and citation networks.
Grounds a coding agent's architecture decisions in real arXiv prior art before it builds something new. Reads arXiv category-wise via real HTTP fetch, spawns parallel isolated reads across the papers found, scores/clusters them, then converges on ONE recommended path with citations, a first step, and known prior-art pitfalls to avoid. Use on /neuroarxiv, before designing non-trivial architecture, algorithms, ML/systems techniques, or protocols, or when the user asks "has anyone solved this", "what's the state of the art", or "am I about to rebuild something that already exists". Skip for trivial CRUD, glue code, or closed phrasing ("just", "quick", "standard"). Full pre-flight gate is in the skill body.
Integrate and use Niko Table (shadcn-compatible data tables with TanStack Table) and the editable Data Grid. Use when the user mentions Niko Table, niko-table.com, data table with shadcn, TanStack Table, building a sortable or filterable table, faceted filters, advanced filter menu, row or column drag-and-drop, server-side pagination, server-side grid, infinite scroll, Drizzle ORM, nuqs, URL state, shareable or bookmarkable table, row expansion, tree table, row selection, editable spreadsheet grid, useDataGrid, DataGrid cell editors, clipboard/fill handle, useGridChanges, or any shadcn-compatible data grid — even if they don't name Niko Table explicitly. Prefer this skill whenever the task involves a data table or editable grid in a React/shadcn project so the model follows Niko Table structure, imports, and patterns.
Odoo 17.0 full-stack development skill for AI agents. Covers backend module architecture, ORM models and fields, API decorators, recordset operations, actions, performance patterns, testing, and built-in mixins. Use when building, extending, or debugging Odoo 17.0 custom modules. Triggers on tasks involving __manifest__.py, models.Model, @api decorators, XML views, ir.actions, ir.cron, mail.thread, or any Odoo 17.0 framework pattern.
Okteto Preview Environment skill. Use when someone wants a live, shareable environment for a branch or pull request — "deploy a preview for this PR", "give me a URL for branch X", posting a preview link back to a PR or thread — or when setting up PR preview automation in CI (GitHub Actions or GitLab). Use alongside the `okteto` skill: for editing, syncing, and iterating on code, defer to that skill's dev environments. Requires an Okteto context; previews are an Okteto Platform feature.
Query OpenAlex REST API for 250M+ scholarly works, authors, institutions, journals, concepts. Search by keyword, author, DOI, ORCID, or ID; filter by year, OA, citations, field; retrieve citations, references, author disambiguation. Free, no auth. For PubMed use pubmed-database; preprints use biorxiv-database.
Compatibility router for the shared optimization knowledge base and the language-specific optimization catalog skills. Use when: (1) selecting which optimization catalog skill to load, (2) the implementation language is not fixed yet, (3) a workflow still references the legacy optimization-catalog skill name, (4) deciding whether a finding is shared or language-specific, (5) updating the generalized knowledge-base structure.
Search and sort results in paper notes in a local Obsidian Vault by title, author, keyword, tag, field, or arXiv ID. Use when the user asks to find, filter, or revisit papers already captured in local notes; do not use for downloading new papers.
Structured manuscript/grant review with checklist-based evaluation. Use when writing formal peer reviews with specific criteria methodology assessment, statistical validity, reporting standards compliance (CONSORT/STROBE), and constructive feedback. Best for actual review writing, manuscript revision. For evaluating claims/evidence quality use scientific-critical-thinking; for quantitative scoring frameworks use scholar-evaluation.
Optimize software performance through user-journey baselines, profiling, representative benchmarks, controlled experiments, and field verification. Use when asked to make an application, service, workflow, or hot path faster or more efficient, or to build the measurement loop first. Do not use for a reliability incident, capacity forecast, generic product A/B test, or operating a named telemetry or load-testing tool without an optimization objective.
A multi-layer data-flow tracing tool from PHP web routes to sinks. Given a user-specified route, it traces from the handler to the final sensitive operation point and outputs layered evidence, parameter and variable tracing, and controllability analysis (without drawing vulnerability conclusions).
Guides use of Pinecone, a managed serverless vector database, through its Python client and the LangChain and LlamaIndex integrations. Covers creating indexes, upserting and querying vectors, metadata filtering, namespaces, hybrid dense and sparse search, index management, and deleting vectors. Use when building a production RAG system on a hosted vector store, adding semantic search or recommendations without running infrastructure, isolating per-user or per-tenant data with namespaces, combining dense and sparse vectors in one query, or filtering results by metadata. Do not use for self-hosted or local stores (Chroma, Weaviate) or offline similarity search (FAISS).
Genre pack: bold, joyful consumer design — giant display type, thick ink outlines, flat vibrant color fields, sticker-like shapes, chunky buttons. For D2C brands, consumer apps, events, education, and youth-facing products. Fun without looking childish: the structure stays strict, only the attitude is loud. Stack on the design-taste core skill.
Use when: you need to validate product optimization results, make data-driven A/B experiment decisions, or assess the causal impact of feature changes. Do NOT use when: changes cannot be measured quantitatively, the sample size is insufficient, or strict statistical validation is not needed
Starting from data phenomena, generate actionable product decision recommendations and output them as a visualized HTML report. Trigger when the user says "analyze this dataset", "why did this metric drop", "do a retention analysis for me", "what's causing user churn", "see where the funnel drops off", or "give me a data analysis report". Also applies when: the user uploads a CSV/Excel data file and asks for analytical insights; the user provides SQL query results and asks for interpretation; the user asks for metric decomposition, attribution analysis, cohort comparison, or A/B test analysis; the user mentions data analysis keywords such as "metric tree", "funnel", "retention", "conversion rate", "DAU decline", "cohort", or "attribution". Not applicable to: pure BI dashboard building (use xlsx), pure SQL writing (write directly), or pure data cleaning (write a script directly).
Use when: you need to define the feature scope of the first product version, filter MVP features from an existing requirements list, or determine the boundaries of a minimum viable product. Do NOT use when: the product is already launched and needs a full feature set, or you only need to organize requirements without trimming scope
Use when: multiple requirements need ordering, resources are limited and you must decide what to do first, or conflicting requirements need objective scoring. Do NOT use when: there is only one requirement and no ordering is needed, or priorities have already been set by a manager or the business side
Use when: risk screening is needed after the technical solution is complete, risk identification and assessment before launch, or risk control at key project milestones. Do NOT use when: the project is very small and risks are under control, or a complete risk management mechanism already exists and does not need to be built anew
Use when: market research data collection, competitor information gathering, industry sentiment monitoring, compliance policy checks. Do NOT use when: complete data is already available and no supplementation is needed, or only internal data analysis is required without external information
Bloom and glow effects using Three.js UnrealBloomPass with React Three Fiber. Use when implementing glow, bloom, luminance-based effects, selective bloom on specific meshes, or neon/ethereal lighting. Essential for cyberpunk aesthetics, energy effects, magic spells, and UI glow.
Invoke whenever spatial SQL or its execution backend is the decision: PostGIS, DuckDB Spatial, SpatiaLite, ST_* functions, recurring spatial joins, concurrent/growing workloads, or large GeoParquet queries. Covers backend selection, schemas, GiST/BRIN indexes, KNN, geometry versus geography, correctness benchmarks, and EXPLAIN optimization. Use PostGIS for managed concurrent services and embedded engines for bounded local analytics when evidence supports that choice. Use geo-data-engineering for acquisition, conversion, and file-based ETL without spatial SQL.
The full Supabase Management API (108 endpoints) plus a local SQLite cache of orgs, projects, functions, branches, and secret names — powering cross-project queries no live API answers in one call, with Auth Admin lookup, PostgREST schema introspection, and Storage usage rollup on top. Trigger phrases: `supabase auth admin lookup`, `supabase secret name audit`, `supabase branches drift`, `supabase project estate rollup`, `supabase storage usage`, `supabase pgrst schema`, `use supabase`, `run supabase`. Anti-triggers: `supabase start` (use the official supabase CLI), `supabase db push` (official CLI), `supabase gen types` (official CLI), supabase realtime subscribe (WebSocket — out of scope).
Use this skill any time a .pptx file is involved in any way — as input, output, or both. This includes: creating slide decks, pitch decks, or presentations; reading, parsing, or extracting text from any .pptx file (even if the extracted content will be used elsewhere, like in an email or summary); editing, modifying, or updating existing presentations; combining or splitting slide files; working with templates, layouts, speaker notes, or comments. Trigger whenever the user mentions \"deck,\" \"slides,\" \"presentation,\" or references a .pptx filename, regardless of what they plan to do with the content afterward. If a .pptx file needs to be opened, created, or touched, use this skill.
Use this skill any time a .pptx file is involved in any way — as input, output, or both. This includes: creating slide decks, pitch decks, or presentations; reading, parsing, or extracting text from any .pptx file (even if the extracted content will be used elsewhere, like in an email or summary); editing, modifying, or updating existing presentations; combining or splitting slide files; working with templates, layouts, speaker notes, or comments. Trigger whenever the user mentions \"deck,\" \"slides,\" \"presentation,\" or references a .pptx filename, regardless of what they plan to do with the content afterward. If a .pptx file needs to be opened, created, or touched, use this skill.
Write or refresh a visual recap of a GitHub pull request for its reviewers, posted as one PR comment between pstack-recap markers. The recap opens with two sentences on why the change exists and what it changes, then one hand-drawn sketch of the changed mechanism, attached as an image, then one line on what the PR does not prove. Risk, SHAs, reading order, and file areas sit in a collapsed block below. Use after you open a PR or push to one, or when asked to “recap this PR”, “visual recap”, “PR recap”, “diagram this PR”, or “summarize this PR for reviewers”. Skip trivial PRs.
Synthesize sound effects from code (numpy/scipy, no sample libraries), check them by numbers, and mix them frame-accurately under a video or animation with voice ducking and masking checks, then master it with a single linear gain to a loudness target for audio-only or video delivery. Use whenever someone needs sound effects or foley for a code-rendered video, animation, explainer, motion graphic or game prototype; wants to generate a specific sound from code (枪声, 爆炸, 脚步, 音效, 拟音, a laser, a door slam); needs SFX with no third-party sample licences; or needs audio synced to animation events, even if they never say "procedural". Not for composing music or finding recorded samples.
analyze a professor from google scholar, publication lists, personal websites, lab pages, and field-specific bibliographic databases (e.g., DBLP, PubMed, SSRN, PhilPapers, MathSciNet, arXiv, Scopus) to evaluate research strength, mentoring quality, collaboration network, lab resources, research taxonomy, future directions, applicant fit, outreach emails, and interview strategy. designed for students at all levels — PhD applicants, master's students, and undergraduate researchers (capstone/thesis/independent study) — across all academic disciplines. use when the user wants to assess whether a professor or lab is worth applying to, compare advisors, prepare a cold email, find a thesis or capstone advisor, infer future research openings, or build a structured dossier from public academic evidence.
Classifies text with Meta's Prompt Guard, an 86M-parameter model loaded from HuggingFace, into BENIGN, INJECTION or JAILBREAK labels to detect prompt injections and jailbreak attempts in LLM applications. Covers user input filtering, third-party data and RAG document filtering, batch processing, threshold tuning, and sliding-window handling of texts over 512 tokens. Use when screening user prompts for jailbreaks before they reach an LLM. Use when filtering API responses or retrieved RAG documents for embedded instructions. Use when batch-scanning documents for injection. Use when tuning detection thresholds or false positives on security-related queries. Use when running a lightweight CPU or GPU classifier for 8 languages. Not for content moderation such as violence or hate; use LlamaGuard for that.
Reasoning skill for healthcare revenue cycle denial management. Use when the user needs to appeal a claim denial, draft a payer-ready denial appeal letter, analyze a denial reason code, or determine whether a denied claim should be overturned, pended, or accepted. Triggers include \"appeal a claim denial\", \"draft denial appeal\", \"write appeal letter\", \"CO-50 denial\", \"medical necessity denial\", \"CARC code appeal\", \"RARC code\", \"overturn denial\", \"prior auth denial\", \"timely filing denial\", \"denial workup\", \"payer denial response\", \"claim denial overturn\", \"CO-96 denial\", \"CO-97 denial\", \"denial letter\", \"denial management\", \"remittance advice denial\", \"835 denial\", \"authorization required denial\", \"ERISA denial\", \"Medicaid managed care denial\", \"visit limit denial\", \"benefit limit denial\", \"inpatient denial\", \"observation downgrade\".
Use for hands-on Pulumi CLI work: running deployments, fixing broken stacks, and managing infrastructure state. Handles: recovering from stuck or interrupted `pulumi up` with pending operations, cleaning orphaned resources from state after out-of-band cloud deletions, protecting critical resources from accidental `pulumi destroy`, moving resources between stacks without recreating them, targeting specific resources during deployment, migrating between backends (local file to Pulumi Cloud, S3), stack lifecycle management, state export/import/repair, CI/CD pipeline setup, and importing existing cloud resources. Use this skill — not the language-specific Pulumi skills — whenever the user's question is about operating, troubleshooting, or recovering Pulumi infrastructure rather than writing program code.
Guidance for applying Microsoft Purview data security and compliance across Microsoft 365 workloads (Exchange, SharePoint, OneDrive, Teams) - coordinating labels, DLP, retention, and Copilot data protection per workload with each workload's distinct behaviour. WHEN: Purview for Microsoft 365, protect SharePoint and Teams data, M365 compliance, labels and DLP across Office, retention for Exchange, Teams data security, OneDrive governance, Microsoft 365 data protection, per-workload Purview.
Builds pytest infrastructure for a Python package — layout and config, coverage gates with fail_under, Hypothesis property-based tests, mutation testing to catch weak or agent-gamed tests, multi-version runs via tox, nox, or uv. Use for 'set up pytest', 'add coverage', 'test on multiple Python versions'. Not for writing tests for one file, fixing a single failing test, or CI workflow YAML.
Adds PyTorch FSDP2 (fully_shard) to training scripts with correct init, sharding, mixed precision/offload config, and distributed checkpointing. Use when models exceed single-GPU memory or when you need DTensor-based sharding with DeviceMesh.
Chaos engineering principles, controlled failure injection, resilience testing, and system recovery validation. Use when testing distributed systems, building confidence in fault tolerance, or validating disaster recovery.
Orchestration skill for enterprise integration testing across SAP, middleware, WMS, and backend systems. Covers E2E enterprise flows, SAP-specific patterns (RFC, BAPI, IDoc, OData, Fiori), cross-system data validation, and enterprise quality gates.
Apply the Holistic Testing Model evolved with PACTS (Proactive, Autonomous, Collaborative, Targeted, Structured) principles. Use when designing comprehensive test strategies for Classical, AI-assisted, Agent based, or Agentic Systems building quality into the team, or implementing whole-team quality practices.
Test quality validation through mutation testing, assessing test suite effectiveness by introducing code mutations and measuring kill rate. Use when evaluating test quality, identifying weak tests, or proving tests actually catch bugs.
Apply XP practices including pair programming, ensemble programming, continuous integration, and sustainable pace. Use when implementing agile development practices, improving team collaboration, or adopting technical excellence practices.
Search the vault using QMD semantic search. Use PROACTIVELY before reading files. Preference order: (1) MCP tools — mcp__qmd__query, mcp__qmd__get, mcp__qmd__multi_get, mcp__qmd__status — if they appear in your tool menu, use them first; (2) CLI `qmd --index <name> ...` as fallback; (3) Grep/Glob only when QMD is not installed. Trigger proactively for past decisions, incidents, people, meetings, architecture, patterns, or any vault content — and after creating/editing notes to check for duplicates and related content.
Use when a user wants an AI assistant or coding agent to provide setup guidance, evaluation, review, improvement, recommendation follow-up, or paired skill/CLI update help for quality management of a project/entity or one of its components/areas. Trigger for requests about quality factors, characteristics, attributes, criteria, areas, factors, requirements, improving a quality factor such as security/reliability/usability, reviewing a QUALITY.md model or evaluation result, evaluating a root area against quality criteria, applying or handing off recommendations, updating the /quality stack, or authoring/improving a QUALITY.md file.
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.