Avoid AI Writing (Claude Skill) — Installieren & Live-Demo

Warum nutzen

Hauptfunktionen

21 categories of AI-tell patterns documented and detected
Auto-rewrite mode with a 'show diff' preview
Style controls (formal vs casual)
Tracks patterns over time so you can see your own AI drift
Compatible with Claude Code, OpenClaw, Hermes

Live-Demo

In der Praxis

avoid-ai-writing-skill.replay ▶ bereit

0/0

Installieren

Wählen Sie Ihren Client

~/Library/Application Support/Claude/claude_desktop_config.json · Windows: %APPDATA%\Claude\claude_desktop_config.json

{
  "mcpServers": {
    "avoid-ai-writing-skill": {
      "command": "git",
      "args": [
        "clone",
        "https://github.com/conorbronsdon/avoid-ai-writing",
        "~/.claude/skills/avoid-ai-writing"
      ],
      "_inferred": true
    }
  }
}

Öffne Claude Desktop → Settings → Developer → Edit Config. Nach dem Speichern neu starten.

~/.cursor/mcp.json · .cursor/mcp.json

{
  "mcpServers": {
    "avoid-ai-writing-skill": {
      "command": "git",
      "args": [
        "clone",
        "https://github.com/conorbronsdon/avoid-ai-writing",
        "~/.claude/skills/avoid-ai-writing"
      ],
      "_inferred": true
    }
  }
}

Cursor nutzt das gleiche mcpServers-Schema wie Claude Desktop. Projektkonfiguration schlägt die globale.

VS Code → Cline → MCP Servers → Edit

{
  "mcpServers": {
    "avoid-ai-writing-skill": {
      "command": "git",
      "args": [
        "clone",
        "https://github.com/conorbronsdon/avoid-ai-writing",
        "~/.claude/skills/avoid-ai-writing"
      ],
      "_inferred": true
    }
  }
}

Klicken Sie auf das MCP-Servers-Symbol in der Cline-Seitenleiste, dann "Edit Configuration".

~/.codeium/windsurf/mcp_config.json

{
  "mcpServers": {
    "avoid-ai-writing-skill": {
      "command": "git",
      "args": [
        "clone",
        "https://github.com/conorbronsdon/avoid-ai-writing",
        "~/.claude/skills/avoid-ai-writing"
      ],
      "_inferred": true
    }
  }
}

Gleiche Struktur wie Claude Desktop. Windsurf neu starten zum Übernehmen.

~/.continue/config.json

{
  "mcpServers": [
    {
      "name": "avoid-ai-writing-skill",
      "command": "git",
      "args": [
        "clone",
        "https://github.com/conorbronsdon/avoid-ai-writing",
        "~/.claude/skills/avoid-ai-writing"
      ]
    }
  ]
}

Continue nutzt ein Array von Serverobjekten statt einer Map.

~/.config/zed/settings.json

{
  "context_servers": {
    "avoid-ai-writing-skill": {
      "command": {
        "path": "git",
        "args": [
          "clone",
          "https://github.com/conorbronsdon/avoid-ai-writing",
          "~/.claude/skills/avoid-ai-writing"
        ]
      }
    }
  }
}

In context_servers hinzufügen. Zed lädt beim Speichern neu.

claude mcp add avoid-ai-writing-skill -- git clone https://github.com/conorbronsdon/avoid-ai-writing ~/.claude/skills/avoid-ai-writing

Einzeiler. Prüfen mit claude mcp list. Entfernen mit claude mcp remove.

Anwendungsfälle

Praxisnahe Nutzung: Avoid AI Writing

Strip AI tells from a blog post before publishing

👤 Anyone shipping AI-assisted writing ⏱ ~15 min intermediate

Wann einsetzen: Right before you hit publish.

Voraussetzungen

Server/skill installed and authenticated — See repo README

Ablauf

Audit

Run the avoid-ai-writing skill on /drafts/launch-post.md. Show me the offending phrases and counts.✓ Kopiert

→ Inline list of detections
Rewrite

Apply the rewrite in-place but show me a diff first.✓ Kopiert

→ Diff you can accept or reject hunk-by-hunk

Ergebnis: A post that doesn't make readers smell the model.

Fallstricke

Don't run it on someone else's prose — Don't run it on someone else's prose — the rewrite changes voice. Use on your own drafts only.

Kosten & Limits

Was der Betrieb kostet

API-Kontingent: See provider docs for rate limits
Tokens pro Aufruf: Varies by tool
Kosten in €: See repo README for pricing details
Tipp: Cache tool results and avoid repeated identical calls.

Sicherheit

Rechte, Secrets, Reichweite

Credential-Speicherung: Use environment variables; never commit secrets

Datenabfluss: Tool calls go to the provider's API as documented

Fehlerbehebung

Häufige Fehler und Lösungen

skill flags valid em-dashes

Tune the threshold in the skill's config — em-dash usage above 1 per 200 words is the default tell.

Prüfen: Reduce density and re-run

rewrites change meaning

Use the diff mode and reject hunks that drift. The skill's confidence score helps prioritize.

Prüfen: Compare original vs rewrite per paragraph

Alternativen

Avoid AI Writing vs. andere

Alternative	Wann stattdessen	Kompromiss
Generic linters	You want full style-guide enforcement	Vale is broader; this skill is AI-tell-specific

Mehr

Ressourcen

📖 Offizielle README auf GitHub lesen

🐙 Offene Issues ansehen

🔍 Alle 400+ MCP-Server und Skills durchsuchen