YouTube MCP for Claude & ChatGPT
TubeLab is the YouTube MCP Server for Claude and ChatGPT. Real channels, viral outliers, video transcripts and much more - all in your AI chat.
翻訳待ち
43 Claude Code skills for making content that doesn't read like a machine made it — prose editing with an offline linter, prompt engineering for 40+ image/video/music models with optional one-command execution, and orchestrators from research to a published carousel, reel or proposal. Russian-first, English throughout.
翻訳待ち
Loading the full README.
Read the full README ↗45 skills for Claude Code that make content — and refuse to let it read like a machine made it.
Prose editing that strips the tells. Prompt engineering for 40+ image, video and music models, with optional one-command execution against the real APIs. And orchestrators that carry a job the whole way: research question → carousel → published post; raw price list → a proposal styled from the client's own website.
Russian-first, English throughout. Plain markdown, MIT, no required dependencies.
The writing tools, with before-and-after examples in English, Russian, Portuguese and Spanish: mikefluff.com/skills.
The writer linter also runs in the browser, nothing installed and nothing sent
anywhere: paste a text and see the phrases that give a model away, in
English,
Russian,
Spanish and
Portuguese.
The same cold email, before and after cold-email → writer:
Subject: Quick question
Hi Patricia,
I hope this email finds you well! My name is David Kim and I'm the co-founder and CEO of Nimbus, a B2B SaaS startup that's been disrupting the developer-tools space for the past 22 months.
I've been a longtime admirer of your work at Magnet Ventures, and I was incredibly impressed by your recent investment in DevForge. Given the synergies I see between our work and your portfolio, I wanted to reach out personally to introduce myself.
We've been growing rapidly — we now have a number of enterprise customers and significant ARR. […]
Subject: $2.1M ARR in dev tools — fit for Magnet?
Hi Patricia,
Saw the DevForge round — congrats on leading. We're in an adjacent wedge: pre-merge code-review automation for engineering teams of 20-200. $2.1M ARR over 22 months, 12 enterprise customers including Stripe, Datadog, and Cloudflare.
Raising A in Q3. Worth 20 min next week to see if Magnet would be a fit?
Happy to send the deck cold either way.
Best, David Kim Co-founder, Nimbus · nimbus.dev
Every cut is attributable to a rule, not to taste — "I hope this email finds you
well" to banned-patterns.md §Ceremony openers, the 24-word bio to
structure.md §Block 1. The full pair with every delta
is in the repo, along with three more.
There is also an offline linter, so the same rules can gate a commit with no API call and no model in the loop:
$ python3 ~/.claude/skills/writer/scripts/lint.py draft.md
writer-lint: borderline (3 hits)
gate passed: no hard bans.
Hits:
L1:42 [caution] GPT_FILLER "it's important to note"
L2:30 [caution] GPT_FILLER "Let's dive into"
L6:1 [caution] AI_BRIDGE 'Moreover'
Exit codes are hook-shaped, so style-check can run it on staged files and stop
the commit — pre-commit walkthrough.
One skill does the prose work — writer — and eleven others
route through it rather than restating its rules. viral-text knows about hooks
and platforms; it does not know about em-dashes, because writer does. The same
shape holds for media: nine skills build on image-prompt.
That is the whole architecture. 45 skills: 1 base, 22 wrappers, 3 linters, 14 orchestrators, 3 meta.
# Plugin marketplace — inside Claude Code
/plugin marketplace add Mikefluff/skills
/plugin install mikefluff-skills@mikefluff
# Curl — 5 seconds, no deps
curl -fsSL https://raw.githubusercontent.com/Mikefluff/skills/main/install.sh | bash
The plugin route keeps the skills updatable from inside Claude Code
(/plugin marketplace update); the curl route drops them straight into
~/.claude/skills/. Either works — don't run both, or every skill appears twice.
Other methods (npm, Homebrew, Docker, local checkout, pinned version, custom prefix): docs/INSTALL.md.
Skills appear after Claude Code session restart. Discovery is automatic via the name: and description: fields in each skill's frontmatter — no ~/.claude/settings.json edits required.
API keys (for --execute mode — entirely optional): manage via /skills-keys add OPENAI_API_KEY ... inside Claude Code, or copy .env.example to ~/.skills.env. See Managing keys below.
The full doc hub: docs/. Most-used pages:
| Page | What |
|---|---|
| QUICKSTART | 5-minute first run — install → first prose edit → first AI image → first end-to-end |
| USER-GUIDE | The scenarios index — pick what you want to do |
| walkthroughs/ | 20 step-by-step recipes, categorized |
| CONTRIBUTING · CODE_OF_CONDUCT | Adding a skill, reporting a bug, house rules |
| SKILL-INDEX | Every skill by layer / domain / language |
| COMPOSING | Named workflows for chaining skills + data flow + anti-patterns |
| INSTALL | Install methods (curl, npm, brew, Docker, local, pinned) |
| FAQ · TROUBLESHOOTING | When something looks off |
This is the beginning of the README. Read the rest at the original. Read the full README ↗