Daily · 16 May 2026
Top 100 Claude Code Skills in the Wild
Ranked from 100 down to 1. Generated by /lad, illustrated by /iad.
#1
skill-creator
Anthropic's official skill for creating, evaluating, and optimising other skills. Underpins almost every skill in the wild — used to scaffold, test triggering, and tune descriptions.
#2
abc (Mat Siems)
Super Master Agent for the 1G 2026 goal — Apps + Backoffice + Context. Central brain routing across ~50 abc-* sub-skills and tracking progress through Prodgame.
#3
aibo (Cleverfox)
Master Agent for the Cleverfox AIBO platform — the AI-first backoffice that drives personas, assets, media, and workflows. Routes to all per-page and codebase agents.
#4
vidai / vad
Video A Day — full daily YouTube-video pipeline from NotebookLM cinematic to ffmpeg assembly. Mat's flagship media-automation skill with `vidai app`, sheet sync, and uploader.
#5
lad (List A Day)
Generate one fresh Top-100 list per day, cross-checked against ~810 historic lists in MongoDB AIDB.DAILY. Powers the LAD gallery on localhost:24205.
#6
pad (Page A Day v2)
Daily 5-stream context digest — personal, project, industry, AI, world. Builds 365 structured pages per year. /pad-research is the legacy keyword variant.
#7
audio (AUDIO project)
Master agent for everything audio — DJ recordings, TTS, Whisper transcripts, AAD pipeline, and a Next.js gallery on port 24202. Auto-launches the app every invocation.
#8
wbp (Watch-hour Build Pipeline)
YouTube Partner Program watch-hour automation for Siems Production. Daily 10-minute video drops, top channel analysis, generated-video inventory.
#9
agents-md
Generates the AGENTS.md standard file for repos — the agent equivalent of CLAUDE.md. Inspired by agents.md and adopted across the open-source agent ecosystem.
#10
spec (Spec-driven Dev)
Brainstorm → spec → chunked execution with subagents. Inspired by superpowers and get-shit-done. Structured replacement for ad-hoc 'just code it' prompts.
#11
rich-ui
Design and build rich, structured UI responses — cards, forms, dashboards, data visualisations. Inspired by Google's A2UI proposal.
#12
knowledge
Knowledge-graph and memory management for agents. Inspired by Cognee — builds RAG pipelines and persistent agent memory across sessions.
#13
research
Deep research for development decisions — web search, repo analysis, library comparison, ADR-shaped writeups. Inspired by Perplexica and autoresearch.
#14
frontend-design
Create distinctive, production-grade frontend interfaces. Anthropic-published skill that avoids generic AI aesthetics — gradients, drop-shadows, glass-morphism.
#15
nano-banana (Gemini image gen)
Image generation wrapper around Gemini CLI's Nano Banana model. The default 'generate any image' skill — blog featured images, thumbnails, diagrams, illustrations.
#16
abc-jsx
Turn any URL, project, or idea into a single-file React JSX artifact using the ABC app shell — sticky header, collapsible nav + context panel, three switchable views.
#17
abc-architect
Cleverfox-stack-aware system architect that turns intake + sector cards into an ADR + data_model + api_spec triad. Decision-document-shaped output, not Mermaid slop.
#18
abc-proposal
Turn any client brief into a full proposal pack — markdown master, Cleverfox-styled HTML, interactive JSX prototype, print PDF, and follow-up email — in one run.
#19
abc-social (SMM)
Take any concept and produce a 13-slide pitch JSX + AIBO page + Remotion video + outreach email + LinkedIn post + tweet. Full multi-format content pack from a single idea.
#20
abc-diagrams
Editorial-quality SVG diagrams (flowcharts, quadrants, pyramids, timelines) assembled into per-topic repo pages. Brand-aligned alternative to Mermaid slop.
#21
abc-intake
Turn any raw client input — voice memo, transcript, scribbled bullets — into a structured Project Card that every downstream ABC skill consumes.
#22
abc-sector
Classify an ABC Project Card into a sector vertical (charity, care, property, recruitment…) and attach the matching pack: jargon, KPIs, persona patterns, compliance.
#23
abc-ccc
Claude Code Cheat — a 4-hour, 10-POM crash course for teams adopting Claude Code. £500 product built from real production experience shipping 13 sites + bo28 + media pipelines.
#24
abc-cleanup
Codebase cleanup pass — lint, unused imports/deps, theme consistency, file organisation. Applied across ABC repos as a pre-PR routine.
#25
abc-gws (Google Workspace cockpit)
Build the ABC GWS cockpit — multi-tenant Workspace command runner with Fox personas, recipe orchestration, and live audit streaming. 5 sequenced CC prompts.
#26
abc-nextjs
Next.js 16 App Router conventions for ABC — route groups, params as Promise, API routes, config patterns. Reference for every new ABC app.
#27
abc-tailwind
Tailwind v4 + shadcn/ui (base-nova) for ABC. CSS import order, theme tokens, dark mode, CVA pattern, fox-orange accent.
#28
abc-shadcn
Manage shadcn components — add, search, fix, debug, style, compose. Project context, component docs, --preset codes, components.json wiring.
#29
abc-uxd
ABC design system — shadcn-aligned CSS variables, primitives, brand tokens. Canonical token set read before writing the first line of styled JSX.
#30
abc-uxd-audit
Whole-site UX audit across mobile/tablet/desktop — responsiveness, a11y, interaction patterns, visual consistency. Updates design system + storybook + backlog.
#31
abc-ms-context
Master Agent for the CONTEXT pillar (~/context-2026) — Mat's private 1G planning, world context, strategic reference. PRIVATE; never exposes contents externally.
#32
abc-smm
Social Media Manager — given any URL, generates 6 assets: blog post, 13-slide pitch JSX, AIBO page, email, LinkedIn, tweet. Supports `--type` for single-asset runs.
#33
aibo-agent-creator
Scaffold AIBO-specific agents — codebase agents, org agents, task agents. Encodes Cleverfox's per-codebase / per-org / per-task layering.
#34
aibo-app-creator
Scaffold a new Next.js app following AIBO conventions — Next.js 15 App Router, Supabase, Tailwind, shadcn/ui, OrgProvider, ToastProvider.
#35
aibo-brief
Turn the contents of briefs/<vertical>/ into a working /<vertical>/brief page in the AIBO app. Idempotent — re-runs update rather than duplicate.
#36
aibo-diagram
Scaffold a vertical-specific diagrams page in the AIBO app. Mirrors the platform-wide /diagrams page but scoped to one vertical, with SVG conventions and theme tokens.
#37
aibo-pdf
Generate a landscape Cleverfox-branded PDF of AIBO documentation. Orange/black/white branding, intended for client and stakeholder distribution.
#38
aibo-vault
Create and maintain an Obsidian vault mirroring AIBO docs, backlog, team notes, and session history. Markdown-compatible knowledge layer.
#39
ama (Anthropic Managed Agents)
Coach + guide navigator + quiz runner + app launcher for the Claude Managed Agents harness (Anthropic's launched-April-2026 product). Study-friendly skill in ~/APPS/ama.
#40
cac (Claude Architect Certification)
Exam coach + study-guide navigator + quiz runner for the CCA cert. Lives in ~/APPS/cac with domain quizzes and resource indexes.
#41
aip (AI Platform / Onyx)
Master agent for the Onyx-based enterprise AI platform at ~/CF/onyx. Setup, dev ops, debugging, migrations, service health — all via /aip.
#42
pad-research
Legacy keyword-research variant of /pad — NLM web research → blog post + portrait infographic + slide deck → HTML/PDF → GSheet log.
#43
wbp (Watch Build Pipeline)
YPP watch-hour growth automation. Daily 10-minute YouTube video drops, top viewed channel analysis, generated-video inventory tracking.
#44
vidai-yt
YT-PIPELINE — one command full VAD26 workflow: research → notebook → 5 production assets (audio, video, deck, infographic L+P) → metadata → report.
#45
vidai-notebooklm
NotebookLM CLI skill (notebooklm-py package). Manage notebooks, add sources, generate assets, ask questions. Used standalone or as part of /yt.
#46
vidai-showcase
Turn any codebase/repo/URL/screenshot into a 24-second showcase MP4 — analyse → boot locally → Playwright capture → Claude script → Remotion render.
#47
vidai-monitor
Monitor long-running lab processes — build pipelines, data imports, AI generation jobs. Tracks progress of background tasks across the lab.
#48
vidai-qa
Full QA lifecycle — all tests, all features, all pages. Debug failures, add bugs to backlog, fix issues, update status. End-to-end /qa.
#49
vidai-vercel
Vercel operations — sync env vars from .env.local, manage deployments, check status, promote, rollback, configure project settings.
#50
vidai-github
GitHub operations via gh CLI — starred repos, repo analysis, issues, PRs, releases, integration discovery.
#51
vidai-release
Release management — version bumping, changelog generation, release notes, tag creation. Standard release pipeline for VIDAI codebases.
#52
vidai-env / abc-env
Validate .env files, check required variables, compare environments, generate templates. Used pre-deploy across every ABC app.
#53
vidai-perf
Application performance analysis — bundle size, build times, component complexity, database query patterns. Identifies and fixes bottlenecks.
#54
claude-api
Anthropic-published skill — build, debug, optimise Claude API / Anthropic SDK apps. Always includes prompt caching; handles version migrations (4.5→4.6→4.7).
#55
init
Built-in skill — initialise a new CLAUDE.md file with codebase documentation. Shipped with Claude Code; usually the first skill any new repo invokes.
#56
review
Built-in skill — review a pull request. Reads the diff, runs checks, leaves structured comments mirroring senior-engineer review patterns.
#57
security-review
Built-in skill — complete a security review of pending changes on the current branch. OWASP-aligned, focuses on data flow and trust boundaries.
#58
loop
Run a prompt or slash command on a recurring interval. /loop 5m /foo, or omit interval to let the model self-pace. Powers polling and watch-style workflows.
#59
schedule
Create, update, list, or run scheduled remote agents on a cron schedule or once at a specific time. Routines for autonomous tasks.
#60
import
Import data from external sources — JSON files, CSV, API responses, other AgentAI instances. The 'load this data' entry point.
#61
push
Push the current git branch to origin. Pre-push summary (branch, last commit, status) then git push. Universally invoked.
#62
simplify
Review changed code for reuse, quality, and efficiency, then fix any issues found. Anthropic-published — refuses to add things, only removes.
#63
update-config
Configure the Claude Code harness via settings.json — hooks, permissions, env vars. Anthropic-published; required for any 'whenever X happens' automation.
#64
keybindings-help
Customise keyboard shortcuts, rebind keys, add chord bindings, modify ~/.claude/keybindings.json. Anthropic-published.
#65
fewer-permission-prompts
Scan transcripts for common read-only Bash and MCP calls, then add a prioritised allowlist to .claude/settings.json to reduce permission prompts.
#66
reformat
Reformat markdown files — remove blank lines within sections, normalise spacing around headings. Tiny, idempotent housekeeping skill.
#67
persona
Manage persona lifecycle — create, version, clone, compare, bulk operations. Central to AIBO-style multi-character agent ecosystems.
#68
ux (UX audit)
Review the whole site UX flow — mobile, tablet, desktop responsiveness, a11y, interaction patterns. Updates design-system and storybook pages.
#69
superpowers
Open-source meta-skill that adds 'planning, focus, ownership' subagent patterns. One of the most-forked skills in the wild.
#70
get-shit-done
Spec-driven skill encoding the 'brainstorm → spec → execute' loop. Influential prior art that the official /spec skill draws from.
#71
cognee
Knowledge-graph + agent memory framework. Used directly as a skill via /knowledge — manages what agents know and remember across sessions.
#72
Perplexica skill
Open-source Perplexity-like research assistant exposed as a skill. Foundation for the /research family that competes with WebSearch.
#73
autoresearch
Autonomous multi-step research framework — surfaces, synthesises, cites. Powers the deep-research mode in several skills.
#74
A2UI (rich-ui prior art)
Google's Agent-to-User-Interface proposal. The conceptual basis for /rich-ui — structured cards, forms, dashboards rather than plain text.
#75
MCP Inspector skill wrapper
Community skill wrapping the MCP Inspector for debugging in-development MCP servers. Pairs with the /mcp catalogue.
#76
mcp (catalog) skill
Skill that enumerates available MCP servers in the current installation and links to their docs. Useful as a 'what tools do I have' index.
#77
git-flow skill
Encapsulates branching, rebasing, and PR-prep patterns. One of the earliest community skills shared in the Claude Code Discord.
#78
test-runner skill
Runs the project's test suite, parses failures, surfaces them as task list items. Common starting point for new Claude Code adopters.
#79
deploy skill
One-command production deploy with branch + commit + SBOM check. Common in CI/CD-heavy teams that want a single agent entry point.
#80
dbt skill
dbt project navigator — run models, test, docs, lineage. Common in analytics-engineering teams adopting Claude Code.
#81
stripe-build skill
Build Stripe integrations from product specs. Encodes Stripe API patterns, webhook handling, and test-mode/live-mode setup.
#82
supabase-build skill
Supabase project scaffolder with RLS, auth, and edge-function patterns. Often paired with abc-nextjs or aibo-app-creator.
#83
fly-deploy skill
Fly.io deployment helper — fly.toml generation, scaling, secrets, regions. Popular in indie-dev community.
#84
vercel-build skill
Vercel project scaffolder distinct from vidai-vercel — covers Next.js, Edge Config, KV, Postgres, and AI SDK patterns.
#85
obsidian-vault skill (Mat)
Mat's Obsidian-vault helper — feeds notes from CC sessions into a vault that supports daily, weekly, and topic-based views.
#86
vault-search skill
Cross-vault semantic search using a local embedding model. Used heavily in /aibo-vault and ABC ccc-vault workflows.
#87
puppeteer-flow skill
Encapsulates common Puppeteer / Playwright patterns — login flows, scraping with pagination, file download verification.
#88
linear-pm skill
Project-management skill on top of the Linear MCP — sprints, milestones, dependency mapping, async standups. Used in cross-functional teams.
#89
notion-publish skill
Publish output from any skill into a Notion page or database. Used as the 'persist this somewhere shareable' step in many pipelines.
#90
obsidian-publish skill
Publish skill outputs into an Obsidian vault. Equivalent to notion-publish but for the local/Obsidian-first crowd.
#91
remotion-render skill
Programmatic video rendering with Remotion. Backs /abc-social and /vidai-showcase for the actual MP4 output step.
#92
tts-elevenlabs skill
Text-to-speech wrapper for ElevenLabs. Sits underneath /audio and /vidai-yt as the narration backbone.
#93
whisper-transcribe skill
Local Whisper transcription with VTT output. Used by /audio transcribe and the AAD pipeline; supports diarisation.
#94
babysit-prs skill
Polls open PRs and surfaces stalled ones, CI failures, and review-time anomalies. Often paired with /loop for continuous oversight.
#95
incident-response skill
Drives an on-call response — pull observability data, correlate logs, draft a postmortem skeleton. Tight pairing with Sentry MCP.
#96
doc-gen skill
Generate API and project documentation from source code. Mintlify-flavoured output; often paired with /mintlify-publish.
#97
i18n skill
Internationalisation pipeline — extract strings, sync translations, validate locale fallback. Crucial for the 13-site ABC ecosystem.
#98
perf-budget skill
Enforce performance budgets in CI — bundle size deltas, LCP/CLS thresholds, query plan regressions. Built on top of vidai-perf patterns.
#99
skill-explorer (gallery)
Browse and search available skills locally. The on-disk counterpart to mcpservers.org for skill discovery; usually paired with skill-creator.
#100
abc (alias /1g, /2026)
The umbrella orchestrator that fronts the entire ABC family. Routes by intent across 50+ abc-* sub-skills. Most-typed command in Mat's workflow.