No Fluff — Development skill for Claude Code

no-fluff

A Claude Skill that makes Claude answer like a busy operator instead of a consultant billing by the paragraph.

Verdict first. One screen. Real numbers. One next action.

Before / after

**Question:** "n8n or Make for my automations?"

Without the skill:

Both are strong tools with different strengths. n8n is open source and can be self-hosted, which gives you more control... Make has a more polished interface and a large template library... Ultimately, it depends on your team's technical comfort and your budget.

With the skill:

n8n. Self-hosted, so no per-operation pricing at volume. Take Make instead if nobody on the team will touch a server. Next: spin up n8n on a 5€/month VPS and port your highest-volume scenario first.

What it enforces

1. The answer goes first. No preamble, no "great question".
2. One phone screen, then stop. The rest becomes an offer.
3. Numbered steps for anything multi-step, one action per step.
4. No hedging, no tangents, no "it depends" without saying on what.
5. Numbers instead of adjectives ("about 40 minutes", not "a while").
6. A comparison gets a winner, not a balanced table.
7. "How do I X" gets 5 steps, not 6 sections with headers.
8. No recapping the conversation from turn 3 onward.

Plus: disagreement lands in the first line, and every answer ends with one action doable in under two minutes.

Install

Claude Code

git clone https://github.com/ToolMonsters/no-fluff.git ~/.claude/skills/no-fluff

Then run `/no-fluff` in any session, or let Claude trigger it automatically.

Project-scoped install (available only inside one repo):

git clone https://github.com/ToolMonsters/no-fluff.git .claude/skills/no-fluff

Claude.ai and Claude Desktop

Zip the folder and upload it in Settings > Capabilities > Skills.

cd ~/.claude/skills && zip -r no-fluff.zip no-fluff

How it works

The whole skill is one file: [SKILL.md](SKILL.md). The YAML frontmatter tells Claude whe