AI Agent Hub
Back to skills
Drunk Mode icon

Drunk Mode

Life Service Updated 2026.08.30

Paste the following prompt into your AI chat to install this skill:

Install @user_8843d1eb/drunk according to https://skillhub.cn/install/skillhub.md.

About this skill

Problem

When AI replies are too polished, they can feel logical but overly manual-like. Drunk Mode addresses this expression layer: it does not replace weather, code, or search capabilities, but adds a lightweight “drunk” persona to ordinary replies so they read more like casual chat while staying legible and bounded.

How It Works

  • State control: toggle it with “enable drunk mode” or “disable drunk mode”, and set a level with enable drunk mode [1/3/5].
  • Three tiers: Level 1 is relaxed and natural, Level 3 is buzzed, Level 5 is visibly chaotic; higher levels use more jumping thought and pauses.
  • Style constraints: metaphors, self-interruptions, and fragmented sentences are allowed, but repetition, typos, and length are limited to avoid gibberish.
  • Runtime behavior: it is a pure instruction-based skill with no external API; it may generate drunk_state.json at runtime to track state.

Boundaries

It fits chat, light Q&A, and lifestyle conversation. It is not suitable for formal reports, legal or medical advice, or tasks requiring strict formatting. Treat the “drunk” style as a presentation layer, not a source of facts; important information still needs to be clear, accurate, and verifiable.

Use Cases

  • When explaining Dalian weather to a teammate, turn a plain weather answer into a relaxed chat tone.
  • In a group chat, answer whether today is suitable for running with metaphors and light pauses.
  • Test chat persona styles by toggling different drunk-mode levels with enable/disable commands.
  • When drafting lifestyle microcopy, turn a flat list into a warmer, slightly emotional tone.

Best For

  • Chatbot demo engineers who need to showcase different conversational persona effects.
  • Social media copywriters who need to turn factual information into a relaxed tone.
  • QA testers who need to verify level switching and output constraints for conversational styles.
  • Product managers building personal assistants who need friendlier expression for lifestyle Q&A.