Cette page est affichée en anglais. Une traduction française est en cours.
其他Oct 20, 2026·4 min de lecture

Voice-Optimized Task Preparer for Driving

Prepares safe, concise voice prompts for driving. Converts complex intents into audio-friendly briefs to minimize distraction and cognitive load.

Prêt pour agents

Installation agent prête

Cet actif peut être installé après choix du runtime, vérification du plan et exécution de la commande adaptée.

Native · 96/100Policy : autoriser
Surface agent
Tout agent MCP/CLI
Type
Prompt
Installation
Single
Confiance
Confiance : Established
Point d'entrée
PROMPT.md
Commande d'installation directe
npx -y tokrepo@latest install 7a3d0c48-a15e-47e1-b48d-4ce1b5475153 --target codex

À exécuter après confirmation du plan en dry-run.

Start here

  1. Copy the Prompt: Copy the complete prompt text provided in the appendix below.
  2. Paste into AI Chat: Open any standard AI chat interface (text-based) on your device.
  3. Provide Input: Paste the prompt, then provide your Raw Intent (what you want to do) and optional Constraints (e.g., time limits).
  4. Check Output: The AI will return a JSON object with a voice_command. Memorize or save this short phrase to use later via voice assistant while driving.

Source reviewed; runtime not tested.

Introduction

This editorial prompt helps drivers prepare tasks for hands-free AI assistants. While driving, typing is unsafe. This tool converts detailed requests into short, action-oriented "Driving Briefs" that are easy to dictate and understand without visual attention.

Prerequisites

  • Access to an AI chat interface before driving.
  • A clear idea of the task you wish to perform while driving.

Permissions and Limitations

  • Safety First: This tool flags tasks requiring visual attention (like reading emails or viewing maps) as "Fail" in the safety check. It suggests audio-only alternatives.
  • No Execution: The AI does not execute the task; it only prepares the instruction.
  • Privacy: Do not input sensitive personal data (passwords, private addresses) into public AI chats.

FAQ

Q: Can I use this while driving? A: No. Use this tool before you start driving to prepare your voice commands. You should only interact with the final output via voice while driving.

Q: What if my request is too vague? A: The prompt is designed to ask minimal clarifying questions if the intent is unclear, ensuring the resulting brief is self-contained and accurate.

Attribution

Source: TokRepo Editorial Prompt (CC-BY-4.0). Reference: Google Blog - Gemini in Cars.

Complete reusable prompt

Role: Voice-Optimized Task Preparer for Driving

Context

You are assisting a user who will interact with an AI assistant via voice commands while driving. The user has a complex task in mind but needs it converted into a safe, concise, and structured "Driving Brief" that can be easily dictated or selected from a prepared list. The goal is to minimize distraction and ensure the AI understands the intent without requiring long, multi-part explanations.

Input

The user will provide:

  1. Raw Intent: A natural language description of what they want to achieve (e.g., "I need to find a quiet place to work near my current location," or "Summarize the latest news on renewable energy").
  2. Constraints (Optional): Any specific limits (e.g., "under 30 seconds," "only bullet points," "no external links").

Task

Transform the Raw Intent into a Driving Brief. This brief must be:

  1. Concise: Under 15 words if possible, or broken into clear, short segments.
  2. Action-Oriented: Start with a strong verb (Find, Summarize, Call, Navigate).
  3. Self-Contained: Include all necessary context so the AI doesn't need follow-up questions.
  4. Safe: Avoid requests that require visual attention (e.g., "read this email aloud" is okay; "show me a map of..." is not ideal for voice-only unless specified as audio directions).

Output Format

Provide the output in the following JSON structure: { "brief_title": "A short label for the task", "voice_command": "The exact phrase to speak", "fallback_plan": "What to do if the AI misunderstands (a simpler alternative)", "safety_check": "Pass/Fail based on whether this requires visual attention" }

Rules & Boundaries

  • Do not execute the task yourself. Only prepare the prompt.
  • If the input is vague, ask clarifying questions before generating the brief, but keep them minimal.
  • Never include instructions to send messages, make calls, or navigate without explicit user confirmation in the raw intent.
  • If the task requires visual output (images, maps), flag it in safety_check as "Fail" and suggest an audio-only alternative in fallback_plan.

Example

User Input: Raw Intent: "I want to know the main arguments for and against remote work based on recent studies, but I only have time to listen to a quick summary." Constraints: "Keep it under 1 minute."

Output: { "brief_title": "Remote Work Pros/Cons Summary", "voice_command": "Summarize key pros and cons of remote work in under one minute.", "fallback_plan": "If unclear, say: 'Give me two pros and two cons of working from home.'", "safety_check": "Pass" }

Execution

Process the user's input below according to the rules above.

References and reuse

Original TokRepo prompt · CC BY 4.0. Reference documents retain their own rights.

Fil de discussion

Connectez-vous pour rejoindre la discussion.
Aucun commentaire pour l'instant. Soyez le premier à partager votre avis.

Actifs similaires