Findings & checks · 0 flagged
Every scanned point with the score it earned and what moved between them.
First recorded scan — no prior version to compare against.
The primary manifest — the file an agent reads to learn what this artifact does.
<div align="center">
🤖 SkynetLabs — All Claude Code Skills
44 production-grade [Claude Code](https://claude.com/claude-code) skills for content & video, SEO/AEO, client delivery, code review, planning, and token efficiency — battle-tested on real agency work, scrubbed clean, and free to use.
Install · Catalog · How skills work · Contributing · Credits
</div>
✨ What is this?
A curated, ready-to-install library of Agent Skills for Claude Code. Each skill is a folder with a SKILL.md that Claude auto-loads when your request matches it — turning Claude Code into a specialist for whatever you're doing: cutting a vertical reel, writing an Upwork proposal, auditing a diff five ways, onboarding a client repo, or just talking in 75%-fewer tokens.
Everything here ran in a real one-person agency before it was cleaned up for the public. No client data, no keys, no personal info — just the reusable machinery.
🚀 Install
Option A — Plugin marketplace (recommended)
In Claude Code:
/plugin marketplace add waseemnasir2k26/skynetlabs-all-claude-code
/plugin install skynetlabs-all-claude-code@skynetlabs-skillsThat's it — all 44 skills register at once. Update later with /plugin marketplace update skynetlabs-skills.
Option B — Manual (clone + install script)
git clone https://github.com/waseemnasir2k26/skynetlabs-all-claude-code.git
cd skynetlabs-all-claude-code
# macOS / Linux
bash scripts/install.sh
# Windows (PowerShell)
powershell -ExecutionPolicy Bypass -File scripts/install.ps1The installer symlinks (or copies) each skill into ~/.claude/skills/ so they're available in every project. Re-run after git pull to get updates. Want only some? Pass skill names: bash scripts/install.sh caveman humanizer critic-loop.
Option C — Cherry-pick one skill
Copy any single skill folder straight into your skills directory:
cp -r skills/token-and-comms/caveman ~/.claude/skills/caveman # global
# or, per-project:
cp -r skills/token-and-comms/caveman .claude/skills/cavemanClaude detects new skills live — no restart needed.
📚 Catalog
<!-- BEGIN CATALOG (auto-generated by scripts/gen_catalog.py) -->
44 skills across 7 categories.
🎬 Content & Reels
| Skill | What it does |
|---|---|
aeo-daily | Local AEO daily content engine — multi-channel daily content pack from one command. |
cinematic-reel | One raw landscape/cinematic clip → polished vertical 9:16 cinematic reel (grade, beat-snapped title cards, optional VO + ducked BGM) in one ffmpeg pass. |
comment-bait-reel | Build a "Comment X for the links" lead-magnet reel (9:16, ~14s) from your own clips + real tutorial thumbnails. |
demo-promo | Turn a raw screen-recording into a branded landscape promo (hook card, rotating strap, PiP inserts, VO + ducked BGM). |
easton-reel | Talking-head clip → finance-bro / hustle reel with kinetic one-word red-serif punch captions snapped to speech. |
ghl-video-drip | Schedule N uploaded videos as a one-per-day, multi-platform drip via a validated GoHighLevel Social Planner CSV. |
ralph-wiggum-marketer | Autonomous copywriter loop: learns your voice, researches competitors, finds angles, iterates to quality. |
reel-studio | One skill for all short-form reels — five modes (interview, talking-head, tts, kinetic-stoic, motion-graphics). |
social-stack | Build, render & schedule social post packs across LinkedIn, IG, FB, X, Pinterest, YouTube (carousels, cards, drips). |
✍️ SEO, Content & Copy
| Skill | What it does |
|---|---|
citation-monitor | Track brand-mention citations across 4 LLMs (Claude, ChatGPT, Perplexity, Gemini) into a local DB + weekly report. |
humanizer | Strip the tells of AI-generated writing (inflated symbolism, em-dash overuse, rule-of-three, filler) from any text. |
seo-geo | SEO/AEO/GEO optimization suite — content writing, meta tags, schema markup, GEO content. |
upwork-proposal | Win Upwork proposals: mirror the JD, anchor price, weave proof, ask sharp questions, close with a CTA. |
🚀 Client Delivery & Planning
| Skill | What it does |
|---|---|
client-onboard | One command bootstraps a per-client repo (CLAUDE.md + deploy gotchas) so Claude starts pre-loaded. |
critic-loop | LLM-as-judge quality gate: an independent critic grades the artifact and loops build→critique→fix until it passes. |
deploy | Multi-target deploy router (Vercel / Hostinger-ZIP / WordPress / static) with deploy gotchas as hard rules. |
flagship-build | Build a 1-of-1 bespoke flagship demo site (not a template variant) for premium showcase tiers. |
plan-ceo-review | CEO/founder-mode plan review: challenge premises, find the 10-star product, expand or strip scope. |
plan-eng-review | Eng-manager-mode plan review: architecture, data flow, edge cases, tests, perf, observability, rollout. |
review | Pre-landing PR review for SQL safety, LLM trust-boundary violations, and conditional side effects. |
ship | Ship workflow: merge main, run tests, review diff, bump version, update changelog, commit, push, open PR. |
skynet-site | Build a premium client website end-to-end — research, design system, funnel, SEO base, deploy, pitch. |
🔍 Code Quality & Review
| Skill | What it does |
|---|---|
caveman-review | Ultra-compressed code-review comments — one line each: location, problem, fix. |
code-audit-fanout | Five-way parallel code audit — 5 specialist lenses, every finding cites an exact source. |
fanout-ship | Parallel multi-agent ship pattern: 1 task → N isolated PRs → smallest-first merge order. |
qa | Systematically QA a web app (diff-aware / full / quick / regression) with a scored report + screenshots. |
retro | Weekly engineering retrospective from commit history with trend tracking, team-aware breakdown. |
🗜️ Token & Communication
| Skill | What it does |
|---|---|
caveman | Ultra-compressed "caveman" communication mode — cuts token usage ~75% while keeping full technical accuracy. |
caveman-commit | Ultra-compressed commit-message generator (Conventional Commits, subject ≤50 chars). |
caveman-help | Quick-reference card for all caveman modes, skills, and commands. |
compress | Compress natural-language memory files (CLAUDE.md, todos) into caveman format to save input tokens. |
🧰 Meta, Browser & Tools
| Skill | What it does |
|---|---|
browse | Fast headless browser for QA & dogfooding — navigate, interact, assert, screenshot (~100ms/command). |
gstack | Headless-browser QA toolkit bundling browse + qa + ship + review + plan workflows. |
jarvis-secure | One-command Windows security sweep + speed/hardening pass (read-only audit, then optional hardening). |
n8n-skills | Bundle of 7 n8n automation skills — code, expressions, node config, validation, workflow patterns. |
setup-browser-cookies | Import cookies from your real browser into the headless browse session for authed QA. |
simmer | Iterative refinement loop — investigation-first judges read the artifact and propose evidence-based improvements. |
simmer-generator | Generator sub-skill for simmer (dispatched by the orchestrator). |
simmer-judge | Judge sub-skill for simmer — scores the artifact + suggests next-round improvements. |
simmer-judge-board | Judge-board sub-skill for simmer — a panel of judges with deliberation. |
simmer-reflect | Reflect sub-skill for simmer — tracks iteration history and the best candidate. |
simmer-setup | Setup sub-skill for simmer — inspects the artifact and infers an evaluation contract. |
skill-builder | Meta-skill: turn a repeated workflow into a house-style SKILL.md. |
🛒 E-commerce
| Skill | What it does |
|---|---|
etsy-empire | Etsy digital-product operating system — niche launch strategy, catalog scaling, compliance guardrails. |
<!-- END CATALOG -->
The catalog above is generated from each skill's frontmatter byscripts/gen_catalog.py. Run it after adding a skill:python scripts/gen_catalog.py.
🧠 How skills work
A skill is just a folder with a SKILL.md:
skills/<category>/<skill-name>/
└── SKILL.md # YAML frontmatter (name + description) + instructions
└── scripts/, references/, templates/ # optional supporting filesClaude reads the description of every installed skill and auto-loads the matching one when your request fits — or you invoke it directly as /<skill-name>. The frontmatter is tiny:
---
name: humanizer
description: Remove signs of AI-generated writing from text. Use when editing or
reviewing text to make it sound natural and human-written.
---Learn more in the official docs: Skills · Plugins.
🤝 Contributing
PRs welcome — new skills, fixes, better docs. See CONTRIBUTING.md and use the new-skill issue template.
📦 Third-party skills & credits
Some skills here were authored by others and are included with attribution under their original licenses (gstack, browse, simmer, n8n-skills, seo-geo, ralph-wiggum-marketer). Full attribution in CREDITS.md.
📄 License
The skills original to this repo are released under the MIT License. Third-party skills retain their own licenses (see CREDITS.md).
<div align="center"> Built with Claude Code. If this saved you time, a ⭐ helps others find it. </div>