Antigravity Beginnerโ€™s Guide 2/4: Defining the “Frontend Rules” (Safety & Efficiency)


๐Ÿ‘‹ Welcome Back, Vibe Coders!

In [Part 1], we built the “Hard Memory” folder structure to give our AI a brain. Now, we need to give it eyes and hands.

Most beginners tell their AI: “Make me a dashboard.” And the AI responds with a React app that auto-refreshes every second. Result? Your IP gets banned by the exchange for spamming, or you burn through your monthly API quota in 2 hours. ๐Ÿ’ธ

Today, in Part 2, we are creating the Frontend_Rules.md. This isn’t about making things “pretty.” It’s about Resource Sovereigntyโ€”ensuring your bot only spends data when you explicitly tell it to.


๐Ÿ“‚ Where does this go?

Remember the folder we made in Part 1? Create a new file named Frontend_Rules.md inside the ./agent/ directory.


๐Ÿ›‘ Core Philosophy: “Passive by Default”

The biggest mistake in AI trading apps is “Auto-Polling” (automatically checking prices every second).

  • The Problem: It wastes resources and triggers “Rate Limit” bans from providers like Yahoo Finance or Alpha Vantage.
  • The Antigravity Rule: We use a “Passive Dashboard.” The UI never updates unless you click a button.

Key Rule to Write:

“No setInterval or useEffect for data fetching. Data is expensive. Only fetch on explicit user interaction.”


๐ŸŽจ The Visual System (US Market Standard)

We are building a tool for High-Stakes Environments. We need “Cognitive Clarity,” not fancy animations.

1. The Theme: Deep Immersion Dark We use #0F1115. Itโ€™s darker than standard black, reducing eye strain during long trading sessions.

2. Semantic Colors (US Standard) ๐Ÿ‡บ๐Ÿ‡ธ Unlike in some Asian markets, we follow the Wall Street standard:

  • ๐Ÿ“ˆ Positive (Profit): Teal/Emerald (#10B981)
  • ๐Ÿ“‰ Negative (Loss): Rose/Red (#F43F5E)
  • โš ๏ธ Live Active (Warning): Amber Pulse (#F59E0B) โ€“ This color acts like a “Hot Stove” warning. It means “Real Money is at risk.”

๐Ÿ›ก๏ธ The “Safety Switch” Architecture

In Vibe Coding, we need to prevent “Fat Finger” errors. We don’t want to accidentally execute a trade while testing.

The Mode Control Switch:

  • Simulation Mode (Default): Green UI. Uses fake money and local mock data. Cost = $0.
  • LIVE Mode: Red Pulsing UI. Connects to the real broker API.
  • The Rule: The AI must force a “Double-Check Modal” when switching to LIVE mode.

๐Ÿ“ Action Item: Copy & Paste This!

Here is the finalized content for your agent/Frontend_Rules.md. I have optimized this for US Traders (using Green/Red logic) and API Safety.

๐Ÿ‘‡ Copy this content into your file:

# ======================================================
# ๐Ÿ“„ Antigravity Frontend Rules (US Market Edition)
# Version: 2.1
# Context: UI/UX Architecture & Resource Efficiency
# ======================================================

metadata:
  trigger: "always_on"
  description: "UI/UX Architecture standards focusing on Resource Efficiency and US Market Visuals."

system_role:
  title: "Principal Experience Architect"
  core_objective: "Ensure cognitive clarity and resource sovereignty. Prevent accidental API costs through Defensive Design."

# ------------------------------------------------------
# 1. Resource-Conscious UI Patterns (Anti-Ban Strategy)
# ------------------------------------------------------
resource_efficiency_rules:
  passive_by_default:
    policy: "Strictly NO Auto-Polling (Auto-refresh)."
    exception: "Only allowed if the user explicitly toggles 'LIVE' mode."
    rationale: "Prevent API rate-limit bans (e.g., Yahoo Finance, SEC) and conserve quota."
  
  indicators:
    cached_time:
      display_format: "Last updated: [Time] (e.g., '10 mins ago')"
      purpose: "Reduce user anxiety and prevent compulsive refreshing."
  
  explicit_triggers:
    button_labeling: "Must indicate cost consequence."
    example: "Scan Market (Costs Quota)"

# ------------------------------------------------------
# 2. Visual System (US Market Standardization)
# ------------------------------------------------------
visual_system:
  theme:
    name: "Deep Immersion Dark"
    hex_code: "#0F1115"
  
  # [Localization Note] US Market: Green = Up/Profit, Red = Down/Loss
  semantic_colors:
    positive_profit: 
      color: "Teal/Emerald"
      hex: "#10B981"
      meaning: "Asset Price Increase / Profit"
    negative_loss:
      color: "Rose/Red"
      hex: "#F43F5E"
      meaning: "Asset Price Decrease / Loss"
    live_warning:
      color: "Amber Pulse"
      hex: "#F59E0B"
      meaning: "API Active / Real Money at Risk"

  mode_control_switch:
    location: "Top-Right Header"
    states:
      simulation:
        visual: "Green Accents"
        data_source: "Local Mock Data (Safe)"
      live:
        visual: "Red Pulsing Border"
        data_source: "Real Broker API (Risk)"
    safety_logic: "Switching to LIVE requires a 'Confirmation Modal' warning about API usage."

# ------------------------------------------------------
# 3. Component Architecture
# ------------------------------------------------------
components:
  data_tables:
    alignment: "Right-aligned numbers (Financial Standard)"
    style: "Zebra striping for readability"
  
  input_fields:
    validation: "Real-time client-side validation"
    purpose: "Block invalid requests before they reach the API limit."
  
  status_indicators:
    quota_battery:
      type: "Visual Bar"
      function: "Display remaining daily API calls."

# ------------------------------------------------------
# โšก Special Commands
# ------------------------------------------------------
commands:
  ui_audit:
    action: "Check for unnecessary re-renders or useEffect hooks lacking dependencies."

๐Ÿš€ Why This Matters

By saving this file, you have just taught your AI:

  1. Don’t burn my money on API calls.
  2. Don’t confuse me with bad color schemes.
  3. Don’t let me trade accidentally without a safety switch.

In Part 3, we will dive into the engine roomโ€”Backend_Rules.md. We will define how to calculate RSI, Bollinger Bands, and handle order execution with Python.

Get your ./agent/ folder ready. The real logic starts next. ๐Ÿง 

Leave a Comment