AI & ML
AI & ML agent skills cover the machine-learning workflow itself: writing and evaluating prompts, building RAG pipelines, running evals, and wiring up model APIs. Each one is a SKILL.md file your agent loads on demand, so the know-how travels across Claude Code, Cursor, and 60+ agents.
-
bbgnsurftech Bundle Designing Database Schemas 3Use when you need to work with database schema design. This skill provides schema design and migrations with comprehensive guidance and automation. Trigger with phrases like "design schema", "create migration", or "model database".
-
benjaminshoemaker Bundle Feature Technical Spec 2Define technical approach (architecture, integration points, data model) for a feature through guided Q&A and write FEATURE_TECHNICAL_SPEC.md. Use after /feature-spec.
-
forgeyclap Skill Forge N8n 2Forge playbook for building and validating n8n automation workflows. Use when creating, fixing, or validating an n8n workflow — keywords: n8n, workflow, webhook, automation, node, trigger, schedule, cron, retry, error branch, credentials, validate_workflow. Orchestrates the ECC n8n skills and n8n MCP tools and enforces the validation gate before claiming production-ready.
-
forgeyclap Skill Forge RAG 2Forge playbook for AI chatbots and RAG systems. Use when building a chatbot, assistant, RAG, or knowledge-base Q&A over documents — keywords: chatbot, RAG, retrieval, embeddings, vector search, ingestion, system prompt, knowledge base, assistant, hallucination, source-aware, lead capture. Covers retrieval logic, grounding, fallback behavior, and safety.
-
pandazki Skill Pneuma ModenameTODO: Describe what this mode's agent does and when it should activate. Example: "Expert at creating and editing [content type] in Pneuma {{displayName}} Mode. Works in a WYSIWYG environment where the user sees edits live in a browser preview panel."
-
arbgjr Skill Reference IndexerIndexa documentos de referencia para uso no RAG. Extrai texto de PDFs, processa e adiciona ao corpus. Use quando: adicionar documento, buscar referencia, listar docs.
-
tankygranny05 Bundle Pull Claude Code Source CodeDownload, prettify, and archive @anthropic-ai/claude-code npm packages. Use when user asks to "archive claude code", "download claude code", "get latest claude code", "npm pack claude code", or needs to set up a new Claude Code version in ~/swe/archive with prettified cli.js and git history.
-
shinpr-claude-code-workflows Skill Recipe Fullstack Build 2Execute materialized fullstack task files with layer-aware agent routing
-
dmzoneill Skill Review Pr MultiagentReview PR Multi-Agent
-
madappgang Skill Error Recovery 2Handle errors, timeouts, and failures in multi-agent workflows. Use when dealing with external model timeouts, API failures, partial success, user cancellation, or graceful degradation. Trigger keywords - "error", "failure", "timeout", "retry", "fallback", "cancelled", "graceful degradation", "recovery", "partial success".
-
madappgang Skill Claudish Usage 5CRITICAL - Guide for using Claudish CLI ONLY through sub-agents to run Claude Code with OpenRouter models (Grok, GPT-5, Gemini, MiniMax). NEVER run Claudish directly in main context unless user explicitly requests it. Use when user mentions external AI models, Claudish, OpenRouter, or alternative models. Includes mandatory sub-agent delegation patterns, agent selection guide, file-based instructions, and strict rules to prevent context window pollution.
-
forgeyclap Skill Forge Heartbeat 2Detect a dispatched agent that has gone silent mid-run — reads the run's events.jsonl and flags any started-but-not-finished agent that hasn't logged anything for longer than a window (default 10 minutes). Use during multi-agent runs, periodically or whenever a run feels quiet, and as a check forge-doctor can call later.
-
forgeyclap Skill Forge Prediction 2Forge playbook for prediction, sports-data, and analysis systems (incl. Telegram tip delivery). Use when building forecasting, sports betting analysis, odds/value models, or statistical prediction with confidence labels — keywords: prediction, forecast, sports, betting, odds, value bet, model, backtest, statistics, confidence, Telegram tips, bot. Enforces uncertainty labels, backtesting, and NO automatic real-money betting.
-
forgeyclap Skill Forge Agent Report 2Structured completion-report contract every dispatched Boss ends its final message with, plus the forge-bin/forge-report.cjs tool that parses, validates, and ingests it. Use when dispatching any subagent (include the contract requirement in the prompt) and when a dispatched agent's final message comes back (validate/ingest it before treating the work as done).
-
anhvu1107 Bundle Fp TS ErrorsALWAYS use this when the request matches FP TS Errors: Handle errors as values using fp-ts Either and TaskEither for cleaner, more predictable TypeScript code.
-
anhvu1107 Bundle Fp TS PragmaticALWAYS use this when the request matches FP TS Pragmatic: A practical, jargon-free guide to fp-ts functional programming - the 80/20 approach that gets results without the academic overhead.
-
anhvu1107 Bundle Hugging Face EvaluationALWAYS use this when the request matches Hugging Face Evaluation: Add and manage evaluation results in Hugging Face model cards.
-
anhvu1107 Bundle Internal Comms CommunityALWAYS use this when the request matches Internal Comms Community: To write internal communications, use this skill for:
-
anhvu1107 Bundle Brand Guidelines CommunityALWAYS use this when the request matches Brand Guidelines Community: To access Anthropic's official brand identity and style resources, use this skill.
-
anhvu1107 Bundle Codebase Cleanup Tech DebtALWAYS use this when the request matches Codebase Cleanup Tech Debt: You are a technical debt expert specializing in identifying, quantifying, and prioritizing technical debt in software projects.
-
anhvu1107 Bundle Error Diagnostics Error AnalysisALWAYS use this when the request matches Error Diagnostics Error Analysis: You are an expert error analysis specialist with deep expertise in debugging distributed systems, analyzing production incidents, and implementing comprehensive observability solutions.
-
anhvu1107 Bundle Context Management Context RestoreALWAYS use this when the request matches Context Management Context Restore: Use when working with context management context restore
-
anhvu1107 Bundle Documentation Generation Doc GenerateALWAYS use this when the request matches Documentation Generation DOC Generate: You are a documentation expert specializing in creating comprehensive, maintainable documentation from code.
-
anhvu1107 Bundle Performance Testing Review Multi Agent ReviewALWAYS use this when the request matches Performance Testing Review Multi Agent Review: Use when working with performance testing review multi agent review
-
biotender-max Bundle Bio OrchestratorMeta-agent that routes bioinformatics requests to specialised sub-skills. Handles file type detection, analysis planning, report generation, and reproducibility export.
-
bartosh16 Bundle AI AuditAudytuje tekst, artykuł, bloga, landing albo cały korpus treści pod kątem sygnałów "AI slop" — produkcji liniowej z jednego master promptu: powtarzalny szablon, brak głosu autora i doświadczenia z pierwszej ręki, zerowy information gain, wzorce językowe LLM, ślady maszynowej produkcji, treść podporządkowana lejkowi sprzedażowemu. Zwraca SLOP-score 0–100, werdykt, dowody z cytatami i plan naprawy. ZAWSZE używaj tego skilla gdy użytkownik mówi: "zrób audyt AI", "ai-audit", "sprawdź czy to slop", "czy ten blog jest pisany przez AI", "audyt bloga konkurencji", "przeanalizuj te artykuły pod kątem AI", "czy ta treść to content farm", "oceń jakość tych tekstów", "sprawdź czy nasze artykuły nie wyglądają jak generowane", "audyt treści pod Helpful Content", "czy Google to zdegraduje", "sprawdź te URL-e", "przeanalizuj konkurencję contentowo". Używaj też proaktywnie, gdy użytkownik wkleja kilka artykułów lub linki do bloga i pyta o ich jakość, oryginalność, wiarygodność albo o to, czy widać w nich rękę AI — nawet bez s
-
bartosh16 Bundle StylometriaAnalizuje styl konkretnej osoby/marki na podstawie dostarczonego korpusu tekstów (transkrypty YT, posty social, artykuły blogowe, e-booki, transkrypcje podcastów, wklejony tekst). Wynik: 3 artefakty — raport analityczny ze statystykami (n-gramy, TTR, długość zdań, otwarcia/zamknięcia), styleguide (frazy-kotwice, rytm, hooki, do/don't, adaptacja per kanał) oraz execution-ready system prompt dla agentów piszących w tym głosie. ZAWSZE używaj tego skilla gdy użytkownik mówi „zrób stylometrię", „przeanalizuj styl X", „styleguide z tych tekstów", „nauczcz mnie pisać jak Y", „system prompt w stylu Z", „głos marki z tych transkryptów", „voice fingerprint", „idiolekt", „stylowy odcisk palca", „klonuj styl pisania", „analiza języka X", „raport stylu autora". Pierwsza akcja skilla to ZAWSZE pytanie „Masz tekst, czy mam znaleźć?" przez AskUserQuestion. Skill działa na pojedynczym pliku, folderze plików, URL do bloga/kanału YT/profilu social albo wklejonym tekście — minimum użyteczne to ~10 000 słów, optimum 30 000+. Dla
-
bartosh16 Bundle Model BiznesowyProwadzi rozmowę, z której wychodzi gotowy model biznesowy: pełny Business Model Canvas (9 bloków), policzona ekonomika, dwa kanały dotarcia z progami wyłączenia, plan pierwszych 90 dni i plan B z warunkami wyzwalającymi. Zaczyna od wywiadu o kompetencjach, preferencjach, awersjach i ekonomice osobistej, potem stawia 2-3 hipotezy modelu, a przed domknięciem canvasu wymusza research – najlepiej żywy, w postaci rozmów z ludźmi z segmentu, a nie danych pozbieranych z sieci. Output: model-biznesowy.md + plan-b.md. ZAWSZE używaj gdy ktoś mówi „zbuduj mi model biznesowy”, „jak zarabiać na tym, co umiem”, „pomóż ułożyć biznes wokół moich kompetencji”, „business model canvas”, „na czym mam zarabiać”, „mam umiejętności, nie mam pomysłu na biznes”, „jak wycenić to, co robię”, „czy ten model ma sens”, „przelicz mi ten biznes”, „plan B dla mojej firmy”, „którymi kanałami docierać do klientów”, „przebuduj mój model biznesowy”, „ile klientów potrzebuję, żeby to się spinało”. Skill jest doradczy i neutralny – nie sprzedaje
-
jewgah Bundle Init AgentsScaffold a tailored Claude Code agent team into the current repo — detects the stack, proposes agents/commands/hooks, then writes and wires them into .claude/ and CLAUDE.md.
-
jewgah Skill Review DeepDeep multi-agent code review — parallel agents cross-check each other to catch critical bugs
-
miint-sunny Bundle Nai5 Prompting用于 NovelAI Diffusion V5 的画面构思、插画、漫画、OC 与服装设计,以及将已有方案转换或修改为提示词。
-
miint-sunny Bundle Nai5 Prompting 2用于 NovelAI Diffusion V5 的画面构思、插画、漫画、OC 与服装设计,以及将已有方案转换或修改为提示词。
-
miint-sunny Bundle Oc Costume DesignUse when the user wants an original character outfit, a wardrobe, a clothing design direction, or a visual garment redesign. Converting a settled outfit into prompt text belongs to nai5-writing.
-
miint-sunny Bundle Nai5 Comic WritingUse when the user wants NovelAI prompt text for a settled comic or four-panel storyboard, a page or panel export, or a targeted correction to existing comic prompt text.
-
miint-sunny Bundle Nai5 Costume WritingUse when the user wants prompt wording for a settled outfit, a clothing-only fragment, clothing in a complete image prompt, or a targeted correction to clothing prompt text.
-
yulianuzhnenko Skill Pubmed RAG SynthesizerVectorizes PubMed abstracts using PubMedBERT, builds FAISS indices, and instructs agents to synthesize clinical evidence with exact PMID citations.
Frequently asked questions
What are AI & ML agent skills?
AI & ML agent skills cover the machine-learning workflow itself: writing and evaluating prompts, building RAG pipelines, running evals, and wiring up model APIs. Each one is a SKILL.md file your agent loads on demand, so the know-how travels across Claude Code, Cursor, and 60+ agents.
Which AI & ML skills are most installed?
Popular AI & ML skills on SkillMD right now include designing-database-schemas, feature-technical-spec, forge-n8n. Rankings shift as installs change; sort this page by "Most installs" for the live list.
Do AI & ML skills work with Claude Code and Cursor?
Yes. Every skill here ships as a SKILL.md file, an open format that works in Claude Code, Claude.ai, Cursor, Codex, Windsurf, and 60+ other agents. Install one with npx skillmds@latest add <owner>/<name>, or copy the file into your agent's skills directory.