French (Human-Like) for Openclaw

Bridges the gap between textbook-perfect AI French and authentic, messy, human-like communication.

ivangdavila
v1.0.0
Feb 11, 2026
2
1.4k
0

Install & Download

1. ClawHub CLI

The fastest way to install a skill directly from the registry.

npx clawhub@latest install french

2. Manual Installation

Copy the skill folder to one of these locations

Global
~/.openclaw/skills/
Workspace
<project>/skills/

Priority: Workspace > Local > Bundled

3. Prompt Installation

Copy this prompt to OpenClaw to install it automatically.

Help me install french using Clawhub. If Clawhub is not installed, install it first (npm i -g clawhub).

Prefer to download?

Get the raw skill files in a ZIP archive.

What is French (Human-Like)?

The French (Human-Like) skill solves the problem of robotic, overly formal French generated by standard AI models. While many LLMs produce technically correct grammar, they often lack the personality, regional nuances, and casual flow that native speakers use in daily life. This skill provides a specialized framework for Openclaw Skills to adopt the correct register, use natural contractions, and implement ne-dropping to ensure your agent passes the native test.

By prioritizing casual norms like Tu over Vous in the right contexts and integrating expressive fillers such as bref and du coup, this skill ensures that generated text sounds like it was written by a person, not a machine. Whether you are generating social media copy, peer-to-peer messaging, or localized dialogue, this skill provides the linguistic depth required for true fluency.

French (Human-Like) Use Cases

  • Localizing digital marketing content to resonate with a younger French-speaking audience.
  • Communicating casually with colleagues or friends without sounding cold or hostile.
  • Creating realistic dialogue for creative writing or game development set in French-speaking regions.
  • Fine-tuning professional correspondence to avoid the stiff textbook French typical of translation software.
  • Ensuring regional consistency when targeting specific audiences in France, Quebec, or Belgium.

How French (Human-Like) Works

  1. Identifies the intended audience to set the correct register (Casual vs. Formal) and pronoun usage (Tu vs. Vous).
  2. Applies elisions and contractions like Chuis or T'es to mimic spoken rhythm.
  3. Strips the formal ne from negative sentences (e.g., Je sais pas) to align with modern usage.
  4. Injects natural fillers and expressive common phrases based on the emotional context.
  5. Matches the vocabulary and slang to specific regional markers if the location is known.
  6. Performs a native test to ensure the text is sufficiently rough and realistic.

French (Human-Like) Setup

To integrate this skill into your agentic workflow, add the definition to your skill library.

# Initialize the skill within your Openclaw Skills workspace
openclaw install skill-french-human

Configure your agent instructions to prioritize the ruleset provided in the SKILL.md for all French-language outputs.

French (Human-Like) Data Schema & Taxonomy

The skill organizes its linguistic output based on a taxonomy of registers and regional identifiers.

Element Description Examples
Register Determines the level of formality. Casual, Standard, Formal
Pronoun Sets the social relationship. Tu (casual/friends), Vous (formal/strangers)
Regional Tag Adjusts slang and vocabulary. France (Meuf), Quebec (Char), Belgium (Septante)
Elisions Toggles spoken-style contractions. J'suis, T'es, Y'a
Fillers Injects conversational markers. Euh, Bah, Du coup, En fait

French (Human-Like) Advanced Features

  • Context-aware ne-dropping for casual dialogue and messaging.
  • Regional consistency checks to prevent mixing of France, Quebec, and Belgium dialects.
  • Dynamic punctuation adjustment for texting (dropping spaces before question marks) vs. formal writing.
  • Multi-agent support via Openclaw Skills to ensure consistent personality across different chat interfaces.

SKILL.md


Loading

Related Openclaw Skills

METADATA

Github Stars: 0
forks: 0

Featured*