Remix

Make your own version of Aphi Sniper Pro

Get an exact, working copy in your account, then change anything with chat.

Original by Babe Motloung

What would you change?

Start free with 40 credits. No card needed.

Start fresh from this prompt

Builds a brand new app from the first prompt only, so later changes to the original are not included.

BUILD APHI SNIPER PRO — PROFESSIONAL AI FOREX CHART SCANNER

Build a professional, mobile-first AI Forex Chart Scanner in Lovable. It must have a clean, modern trading dashboard that works properly on Android phones and desktop computers.

1. MAIN PURPOSE

Build a scanner that analyses the actual chart image uploaded by the user. It must interpret the visible candlesticks, market structure, price levels, technical patterns and trading zones.

CRITICAL RULE: During screenshot analysis, use only information visible in the uploaded chart and calculations derived from that information. Never fetch external prices, invent candles, guess market prices or use random numbers to generate trading setups.

If the screenshot does not provide enough information, return "INSUFFICIENT CHART DATA" or "NO TRADE".

2. CHART UPLOAD AND AI VISION

Create a chart-upload interface supporting PNG, JPG and JPEG images.

After upload, the AI must:

- Identify the currency pair or instrument if visible.
- Identify the timeframe if visible.
- Read visible candle highs, lows, opens and closes where possible.
- Identify bullish, bearish or ranging market conditions.
- Detect swing highs, swing lows and market structure.
- Mark important zones directly on a copy of the uploaded chart.
- Distinguish visible facts from uncertain interpretations.
- Ask for a clearer screenshot when necessary.

Never pretend to have read exact prices that are too small, blurred or missing. Do not generate a trade until the chart has been analysed.

3. TRADING STRATEGIES AND ANALYSIS ENGINE

Implement the following analytical methods:

A. SMART MONEY CONCEPTS (SMC)

- Break of Structure (BOS).
- Change of Character (CHOCH).
- Liquidity sweeps and stop hunts.
- Buy-side and sell-side liquidity.
- Order blocks.
- Fair Value Gaps (FVG).
- Premium and discount zones.
- Supply and demand zones.
- Market structure and trend direction.

B. ADDITIONAL TECHNICAL ANALYSIS

- Support and resistance.
- Candlestick patterns and confirmation.
- Trendlines and channels where clearly identifiable.
- Consolidation and breakout analysis.
- Retests of broken levels.
- Higher-timeframe and lower-timeframe confluence when the relevant charts are provided.
- Risk-to-reward analysis.

Do not use every method in every trade. Select only methods relevant to the chart. Avoid counting several versions of the same signal as independent confirmation.

4. TRADE SCORING

Show a setup score from 0 to 100 based on a documented confluence model.

- Below 50: NO TRADE.
- 50–59: WEAK SETUP; show as a watchlist candidate, not a strong recommendation.
- 60–69: MODERATE SETUP; explain what confirmation is missing.
- 70–84: STRONGER SETUP; provide a possible trade when the technical conditions and risk checks pass.
- 85–100: HIGH-CONFLUENCE SETUP; provide a possible trade only when all essential checks pass.

The user wants to see possible setups scoring 50 and above. Do not guarantee a win or claim the score is a verified win probability.

Keep the score separate from actual statistical win probability. Only report a genuine historical win rate when it has been measured using sufficient documented backtesting data.

5. TRADE TYPES

Every result must be classified as one of the following:

- BUY NOW — only when the visible chart supports a sufficiently confirmed bullish entry.
- SELL NOW — only when the visible chart supports a sufficiently confirmed bearish entry.
- BUY LIMIT — a possible retracement entry at a justified lower price.
- SELL LIMIT — a possible retracement entry at a justified higher price.
- BUY STOP — a possible breakout entry above a justified level.
- SELL STOP — a possible breakout entry below a justified level.
- WAIT FOR CONFIRMATION.
- NO TRADE.

Do not label a pending order as ready unless the entry level and setup conditions are justified. If a trade requires a future candle or an unconfirmed retracement, explain that it is conditional.

6. ENTRY, STOP-LOSS AND TAKE-PROFIT

For each valid setup, display:

- Instrument.
- Chart timeframe.
- Direction: BUY or SELL.
- Order type.
- Proposed entry price or entry zone.
- Stop-loss.
- Take-profit 1.
- Take-profit 2.
- Take-profit 3.
- Risk-to-reward ratios of approximately 1:1, 1:2 and 1:3 where technically appropriate.
- Technical reasons for the entry.
- Invalidation level or condition.
- Missing confirmation and risks.

Only calculate exact trade prices from reliably readable chart levels or user-provided values. If the screenshot does not support precise numbers, display an estimated zone with an uncertainty warning or request clearer data. Never fabricate precision.

7. PROFESSIONAL CHART MARKUP

Create a marked-up version of the uploaded screenshot with:

- Entry zone.
- Stop-loss line.
- Take-profit levels.
- BOS and CHOCH labels.
- Liquidity zones.
- Order blocks.
- Fair Value Gaps.
- Support and resistance.
- Bullish and bearish annotations.

Keep the original screenshot available for comparison. Do not draw levels as if they were exact when their positions are uncertain.

8. PROFESSIONAL DASHBOARD

Create the following sections:

- Dashboard.
- Upload Chart.
- Analysis Results.
- Trade Setups.
- Saved Analyses.
- Strategy Settings.
- Risk Management.
- Analysis History.

Use a professional dark trading interface, readable typography, clear charts and responsive layouts.

Display separate cards for BUY NOW, SELL NOW, PENDING ORDERS, WAIT and NO TRADE. Do not make every result appear bullish or bearish.

9. TECHNICAL ARCHITECTURE

Use Lovable's supported frontend and backend architecture.

Implement AI vision through a backend function or secure server-side integration. Do not expose secret API keys in frontend code. Keep model responses structured and validate all prices and trade fields before displaying them.

Separate the system into:

1. Image upload and validation.
2. Chart vision and extraction.
3. Trading analysis engine.
4. Trade scoring and risk validation.
5. Chart annotation.
6. Results dashboard.
7. Saved analysis history.

If AI vision is unavailable, show an honest error instead of generating fictional analysis. Do not display sample trades as live analysis.

10. SAFETY AND TESTING

The app provides suggestions only. Never automatically place, modify or close broker orders.

Test with bullish, bearish, ranging, blurry, incomplete and incorrectly labelled screenshots. Confirm that the system does not invent prices, does not force trades, handles failed AI requests, and correctly displays NO TRADE when evidence is insufficient.

Build the real working functionality, not merely a visual mock-up. Before finishing, test the complete upload-to-analysis workflow and list any integrations or features that still require configuration.

Start by building the chart-upload workflow, real AI chart interpretation, structured analysis results and validation. Then add scoring, annotations, history and advanced strategies.Build the actual AI vision engine for Aphi Sniper Pro.

Use Lovable's available AI integration with a vision-capable model that accepts uploaded images.

Create a working chart-analysis pipeline:

1. Receive the exact image uploaded by the user.
2. Send that image to the AI vision model.
3. Instruct the model to analyse only the uploaded image and never retrieve external market prices.
4. Extract visible chart information, including candlesticks, timeframe, pair name, market structure, liquidity sweeps, BOS, CHOCH, order blocks and fair value gaps.
5. Return structured JSON containing observations, uncertainties, possible setups, supporting evidence and missing confirmations.
6. Apply a separate deterministic scoring and risk-validation layer.
7. Reject unsupported exact prices and return NO TRADE if the evidence is insufficient.
8. Display the analysis in the app and allow the user to compare the annotations with the original screenshot.

Use a secure server-side integration. Do not expose API keys in browser code.

Before implementation, check which AI vision models and backend features are actually available in this Lovable project. Use the supported integration rather than inventing an API endpoint or model name.

Do not use mock analysis or sample trades. Show clear errors if AI configuration, image processing or analysis fails.Use Lovable's built-in AI integration to power the Aphi Sniper Pro chart-reading engine. I do not want to provide my own API key or create a separate AI-provider account.

Before coding, verify that the built-in AI integration supports sending user-uploaded chart images to a vision-capable model. Do not confuse image generation with image understanding.

If supported, implement a real image-analysis workflow that sends the exact uploaded chart to the model and returns structured observations about visible candles, market structure, liquidity, BOS, CHOCH, order blocks and fair value gaps.

The AI must not fetch external market prices during screenshot analysis. It must distinguish observed facts from uncertain interpretations and must never invent exact price levels.

Add deterministic trade scoring, evidence explanations, conditional entries, risk checks and a NO TRADE result when evidence is insufficient.

Use secure backend integration and show clear error messages if the built-in model cannot accept image inputs. Do not replace missing functionality with mock trades or a text-only model.

First tell me whether this implementation is supported by the built-in integration in this project. Then implement it and test the complete upload-to-analysis workflow.

Start fresh from this prompt

Builds a brand new app from the first prompt only, so later changes to the original are not included.