explaingit

alexguoinbj/520-skill

0Audience · generalComplexity · 1/5ActiveSetup · easy

TLDR

A Chinese language joke Skill that switches an AI assistant into a flirty love-brain persona when triggered by the number 520, emotional words, or workplace pain points, while still answering the real question.

Mindmap

mindmap
  root((520-skill))
    Inputs
      Number 520
      Emotional words
      Workplace pain points
    Outputs
      Flirty opener
      Real answer
      Cheesy pickup line
      Warm sign off
    Use Cases
      Add humor to AI chats
      Industry themed jokes
      May 20th online valentines
    Tech Stack
      Markdown
      Skill format
      Prompt template

Things people build with this

USE CASE 1

Add a romantic comedic persona to a Claude or ChatGPT assistant for May 20th

USE CASE 2

Generate industry specific cheesy pickup lines that still solve the real question

USE CASE 3

Wrap technical answers in a warm flirty tone for fun internal demos

USE CASE 4

Study how a Skill package combines triggers, response template, and reference files

Tech stack

MarkdownPrompt

Getting it running

Difficulty · easy Time to first run · 5min

Drop the SKILL.md and reference folder into your assistant's skills directory; no code or dependencies to install.

In plain English

This is a Chinese language joke project that adds a romantic persona to an AI assistant. The number 520 in Chinese internet culture sounds like the phrase for I love you and so May 20th has become a kind of online valentines day. The repo packages a Skill, meaning a set of instructions and reference files, that switches the assistant into a flirty so called love brain mode when the user mentions certain trigger words. The skill activates on three kinds of input. The most direct is the literal number 520. The second is emotional vocabulary like valentines day, confession, breakup, exhausted, or meltdown. The third is workplace pain points such as bugs, going live, project schedules, rollbacks, or thesis writing. Once triggered, the assistant is supposed to keep the romantic tone for the entire conversation, even when the topic in the middle is something serious like a technical architecture or a candlestick chart. The response format is strict and has four parts every time. First a one or two sentence flirty opener that compliments or teases the user about the question they just asked. Then a real, no shortcuts answer to the actual problem, including runnable code if code is what was requested. Then a sudden, very specific cheesy pickup line that uses the vocabulary of the user's profession. Then a warm, companion style sign off. To make the profession specific pickup lines land, the repo carries ten reference files under a reference directory, one per industry. They cover programmers and data analysts, product and UI design, finance and investment banking, marketing and growth, legal, students and researchers, medical and biology, operations and HR and project management, content creators and media, and a catch all for other fields. The SKILL.md file holds the core triggers, the response template, and the behavior rules. There are a few explicit rules about what not to do. The assistant should not fall back on cold robotic phrases like I am only an AI, should not slack off on the technical part of the answer, and should not copy paste tired internet memes. The whole thing is positioned as a way to add a bit of warmth and humor to professional conversations rather than as a serious productivity tool.

Copy-paste prompts

Prompt 1
Install the 520-skill into my Claude Skills directory and trigger it with a message containing 520
Prompt 2
Generate a flirty opener plus a real Python answer plus a programmer themed pickup line plus a warm sign off, using the 520-skill template
Prompt 3
Adapt the 520-skill response template to a new industry not in the reference folder, like aviation
Prompt 4
Use 520-skill to answer a real Bug fix question while keeping the love-brain persona for the whole reply
Prompt 5
Translate the 520-skill SKILL.md triggers and rules into English while keeping the four part response shape
Open on GitHub → Explain another repo

Generated 2026-05-22 · Model: sonnet-4-6 · Verify against the repo before relying on details.