Learn

How to use the platform — from first scan to live deployment.

How to use the platform

A step-by-step guide from first scan to live deployment

1

Start a Research Project

“Create a research project for momentum strategies on large-cap equities”Creates project, sets thesis → visible in your Research workspace
2

Scan Signals

“Scan RSI < 30 on AAPL over the last 2 years at 1h resolution”Runs signal scan → returns p-value, win rate, Sharpe, auto-verdict
“Now test that across all mega-caps”Multi-symbol scan → ranked results by significance
“Sweep RSI period from 10 to 20”Parameter sweep → finds optimal period
3

Build a Strategy

“Promote the AAPL RSI signal into a strategy with a trailing stop”Creates strategy with entry + exit cards, attaches to project
“Add a trend gate — only trade when ADX > 25”Adds gate card, shows updated strategy preview
4

Backtest and Iterate

“Run a backtest”Runs backtest → Sharpe, drawdown, win rate, trade breakdown
“Tighten the stop loss to 2% and compare”Clones strategy, modifies, runs side-by-side comparison
“Sweep the RSI threshold from 25 to 35 in steps of 5”Runs 3 backtest variants, ranks by Sharpe → find optimal parameters
4b

Debug and Understand

“Why did it enter trade #3?”Shows bar-by-bar condition evaluation → exactly which conditions fired and when
“Why didn’t it trade between bar 50 and 100?”Shows which conditions blocked entry at each bar in the range
5

Validate

“Run walk-forward analysis”Divides history into windows, tests consistency → WFE score
“Check it on data it hasn’t seen”Holdout validation → training vs holdout comparison
“Is this overfit?”Overfitting assessment → iteration trajectory, parameter sensitivity, trade clustering
6

Paper Trade

“It looks good, promote to paper”Freezes strategy, starts paper session with benchmark metrics
“How’s paper trading going?”Validation report → pass/marginal/fail verdict

The agent speaks your language

Describe what you want in plain English — the agent maps your intent to the right tools and indicators

Signal Research

“Is RSI oversold actually predictive on AAPL?”Runs signal scan with auto-verdict
“Test that across all mega-caps”Multi-symbol scan
“Sweep the period from 10 to 20”Parameter sweep with ranked results

Opening Positions

“Buy when it’s cheap”RSI oversold entry
“Get in when it breaks out”N-bar breakout entry
“Buy the dip”Percent dip entry
“Enter when the trend starts”EMA crossover entry

Closing Positions

“Get out if it drops too much”Fixed stop loss
“Ride winners”Trailing stop
“Take profit at 5%”Fixed take profit
“Exit after 3 days”Time stop

Filtering Trades

“Only trade when trending”ADX trend gate
“Skip volatile markets”Volatility regime gate
“Only trade on weekdays”Time filter gate

Sizing & Risk

“Put in less when things are wild”Volatility overlay
“Risk $5000 per trade”Fixed USD sizing

Debugging & Understanding

“Why did it enter that trade?”Bar-by-bar condition evaluation for that specific trade
“Why didn’t it trade between bar 50 and 100?”Shows which conditions blocked entry at each bar
“What was happening at bar 75?”Full bar context: price, indicators, condition states
“Show me the condition timeline for bars 40–80”Per-bar evaluation of every condition over a range

Validation

“Is this overfit?”Overfitting assessment
“Test it across different time periods”Walk-forward analysis
“Check it on data it hasn’t seen”Holdout validation
“Are these two strategies redundant?”Correlation analysis between strategy return streams

Events & Scheduling

“Only trade around FOMC meetings”Event-gated strategy with FOMC event type
“Pause trading 30 minutes before earnings”Event distance condition with minute-precision offset
“Buy $500 of AAPL every Monday”Calendar schedule entry (weekly DCA)

Portfolios

“Combine my top 3 strategies into a portfolio”Creates portfolio with selected strategies
“Are my strategies diversified?”Correlation matrix across strategy return streams
“Paper trade the whole portfolio”Starts portfolio paper trading session

Deployment

“It looks good, let’s try it live”Promote to paper trading
“How’s paper trading going?”Validation report with verdict

Strategy building blocks

Gate

Blocks trading when conditions aren’t right

Entry

The buy signal

Exit

Closes your position

Overlay

Adjusts position sizing

Gate → Entry → Exit → Overlay — evaluated in order, every bar

Connect Claude

Connect the MCP server to use the full research workflow

Claude Desktop Pro / Max — Connectors

  1. Settings → Connectors
  2. "Add custom connector"
  3. Name: Quawd
  4. Paste the MCP server URL below
  5. Authenticate with Google

Claude Desktop All tiers — settings file

  1. Settings → Developer → Edit Config
  2. Paste and save:
{
  "mcpServers": {
    "quawd": {
      "command": "npx",
      "args": [
        "mcp-remote@latest",
        "https://vibe-trade-mcp-833596808881.us-central1.run.app/"
      ]
    }
  }
}

Requires Node.js. Restart Claude Desktop after saving.

Claude Code CLI

claude mcp add \
  --transport http \
  quawd \
  https://vibe-trade-mcp-833596808881.us-central1.run.app/

Remote MCP Server URL

https://vibe-trade-mcp-833596808881.us-central1.run.app/

Try these

Click any of these to create a new strategy and open Builder with a ready-made prompt.

These work in the web builder. For the full research workflow (signal scans, walk-forward, holdout), connect Claude via MCP above.