Marketing Skills
coreyhaines31/marketingskillsMarketing skills for Claude Code and AI agents — CRO, copywriting, SEO, analytics, and growth engineering.
The skills, agents, and toolkits we rate for working with Claude — ranked, reviewed, and linked to source. Each is open on GitHub.
Marketing skills for Claude Code and AI agents — CRO, copywriting, SEO, analytics, and growth engineering.
Clone any website with one command using AI coding agents — a Next.js / React + Tailwind + shadcn template.
One skill to create them all — from a single /metaskill command, generate a complete AI agent team, a single expert agent, or a custom workflow skill…
Rebuild any object from a single reference photo as a procedural, animation-ready Three.js model — pure code (primitives, geometry, shaders), no mesh…
A senior-engineering workflow for Claude: structured planning, test-first development, self-review, and helper sub-agents.
Turn any website into clean, LLM-ready data — scrape, crawl, and search the web from your agent.
Reduces token usage by forcing terse, "caveman"-style reasoning — less verbose thinking, lower cost.
A collection of skills for research and scientific workflows — literature, data handling, and analysis.
A long-term memory layer that persists context across projects and conversations.
Strips the tell-tale AI tone from writing so drafts read like a person wrote them.
A large community collection of skills, agents, and slash commands for Claude Code.
66 curated, development-focused skills for Claude Code.
Gives Claude a real browser for clicking, typing, and logging in to do browser-based tasks.
An open-source memory layer for AI agents — gives agents persistent, long-term memory so they can store and recall context, facts, and past interacti…
Gives Claude/AI agents the ability to research what happened in the last 30 days across Reddit, X, Hacker News, YouTube, GitHub, blogs, and news sour…
Improves the frontend/UI design quality generated by AI coding agents — helps produce more polished, aesthetically pleasing interfaces instead of gen…
Lets AI agents build a knowledge graph of a codebase and explore relationships, dependencies, architecture, and functionality through natural languag…
Creates a structured code graph that AI coding agents can use to understand large codebases faster and with better context.
A context-compression skill that reduces token usage by intelligently summarizing logs, documents, tool outputs, and large context before passing it …
Discover and install agent skills from the open ecosystem using the Skills CLI (npx skills) — search, vet by install count and source reputation, the…
Official Anthropic skill that walks Claude through context gathering, iterative refinement and reader testing — turning writing into a workflow, not …
Anthropic's skill for building new skills in the right format — a canonical starting point for authoring your own internal skills.
Official skill for frontend design workflows — gets Claude thinking in interface systems, not random UI fragments.
Teaches Claude to unpack, edit and repack Word/DOCX files reliably — practical whenever your workflow touches structured business documents.
Uses one skill to inspect other skills before you trust them — security review as part of the skill workflow.
Pushes Claude toward a persistent operator model that improves its own process instead of just solving isolated tasks.
Teaches Claude to build for the MCP layer — better external integrations, less brittle glue code.
Retrieval-augmented generation patterns for serious agent workflows — retrieval quality often matters more than raw cleverness.
Brings SLOs, dashboards, alerts and tracing logic into the design process — Claude thinking like infra people.
Teaches Claude to care about bottlenecks, not just features — pushes AI coding into a performance mindset.
Operational skill that treats release discipline as part of product quality, not an afterthought.
Makes Claude reason about technical debt, prioritization and long-term maintenance instead of pretending every codebase is clean.
Turns loose ideas into structured action — a simple building block for a daily AI operating system.
Review, synthesis and better judgment — the reflection loop that good systems need alongside execution loops.
Inbox triage workflows — one of the easiest ways to make Claude feel genuinely useful day to day.
Browses a running app, takes screenshots and produces proper documentation with steps and visuals — documents what exists, not what you hoped.
Converts arbitrary docs into reusable skill memory — stop repeating context, start packaging expertise.
A narrow skill for enriching incomplete author-profile data and related metadata — proof that skills are tiny workflow brains.
Lets Claude take real actions across connected apps — actual execution across Gmail, Slack, GitHub, Notion and more.
Composio's own skill-creation workflow for a more ecosystem-specific packaging flow.
Research, outlines, citations and drafting support — structured writing collaboration rather than generic “write me a post”.
Helps Claude organize cluttered local file systems more intelligently — surprisingly useful immediately.
Apollo.io prospecting and lead-enrichment workflows — a strong example of a vertical SaaS action skill.
Front-end artifact generation that pushes Claude toward polished HTML artifact output.
A niche but real productivity multiplier — isolated git worktrees reduce branch chaos in parallel AI-assisted development.
A specialized cloud workflow pack — narrower, infrastructure-aware guidance for AWS work.
A knowledge-networking skill for linking and summarizing related documents — better research synthesis than flat note dumping.
Persistent, project-scoped deep-research memory that survives compaction and keeps findings alive across sessions.
A standalone deep-research repo with direct clone instructions — deep research as its own serious category.
A 42-skill collection spanning UX, marketing, product strategy, sales, operations, virality, code quality and systems architecture.
A live discovery layer for skills that are actually gaining attention across the ecosystem — escape stale lists.
One of the biggest cross-agent collections — portable skills across Claude Code, Codex, Gemini CLI, Cursor and more.
A widely-shared Claude Code config that keeps the agent disciplined — writes tests before claiming success, stops committing broken code, and persist…
Turns a single Claude Code agent into a virtual engineering team — a CEO that pushes back on the idea, an eng manager that plans architecture, a revi…
“Skills for real engineers, not vibe coding” from TypeScript educator Matt Pocock — makes Claude ask sharp questions first (/grill-me), do test-first…
Turns any folder — code, docs, even PDFs and images — into a queryable knowledge graph so Claude queries a map instead of re-reading every file, cutt…
A memory layer for your whole working life — people, companies, meetings, emails — that synthesizes cited answers rather than returning links, wired …
NVIDIA's security scanner for agent skills — point it at a skill before installing and it checks for hidden instructions, prompt injection, data exfi…
A library of 817 structured cybersecurity playbooks — malware analysis, threat hunting, cloud breaches, incident response — each mapped to frameworks…
Turns Claude into a full video production studio — describe a video and the agent researches, scripts, sources or generates visuals, narrates, edits …
ByteDance's multi-agent harness for long, complex jobs — a lead agent splits the task and runs sub-agents in parallel (scrape, analyze, chart) in a s…
Edit videos with Claude Code — drop raw footage in a folder and chat with an agent to get a finished MP4. It transcribes the audio, cuts filler words…
A single CLAUDE.md file to improve Claude Code behavior, derived from Andrej Karpathy's observations on LLM coding pitfalls.