feat: Add Phase 4 advanced metrics and components

- Add advanced metrics dashboard with trade analytics
- Add new trading components (EntryTypeAnalysis, MultiDayPositionTracker, NewsEventTracker, etc.)
- Add strategy mode selector and trend confirmation
- Add risk automation panel and slippage correlation analysis
- Add daily trading plan enhancements with modal components
- Add custom hooks (useApi, useLocalStorage, useAdvancedTradeMetrics)
- Add broker service integration and trading API
- Add test setup and vitest configuration
- Include parquet data files for live market data
- Add comprehensive documentation in docs/ folder
This commit is contained in:
Krikorios
2025-11-27 10:23:58 +02:00
parent b5e2b02cb8
commit 48e60d015f
2019 changed files with 39793 additions and 257 deletions
@@ -0,0 +1,136 @@
# Gold Trading Assistant Product Vision
## 1. What this product is
This project is an **AIpowered trading assistant** for gold (XAU/USD) traders.
It is designed to:
- Help you **prepare** for the trading day (checklists, AI daily plan, news, economic calendar).
- Support your **decisions** during the session (AI analysis, risk sizing, pattern insights, coaching).
- Make it easy to **log** trades and **review** performance (journal, equity curve, analytics, lessons).
The assistant focuses on **workflow, clarity, and discipline** — not on automation or execution.
---
## 2. What this product is NOT
This product is **not**:
- A broker or exchange.
- An orderexecution or autotrading platform.
- A place where real money trades are placed.
All **actual trades** are executed on **external platforms** such as:
- Your broker (e.g., MT5, cTrader, web platform),
- TradingView broker integration,
- Or any other execution venue you choose.
The assistant only helps you **decide**, **document**, and **improve**.
It never sends orders or connects to your broker account.
---
## 3. Core user
The primary user is a **discretionary or semisystematic gold trader** who:
- Trades **intraday or swing** (1m to 1D timeframes).
- Wants **structure and discipline**:
- Clear daily routine,
- Consistent trade logging,
- Datadriven review and improvement.
- Is comfortable executing trades on an external platform but wants a **“trading copilot”** for analysis and decision support.
We design everything for this user first.
---
## 4. Core workflow: Prepare → Decide → Execute elsewhere → Log → Review
The assistant is built around one daily loop:
1. **Prepare (Today tab)**
- Morning checklist (risk, news, economic calendar, levels).
- AIassisted daily trading plan.
- News and economic events relevant for gold.
- Coaching focus for the day (what to watch in your behavior).
2. **Decide (Trade tab)**
- Live gold price and charts.
- AI analysis (bias, key levels, risk context).
- Risk management tools (position sizing, RR, exposure).
- Clear picture of your current and planned trades.
3. **Execute elsewhere (external platform)**
- You place all trades on your broker or charting platform.
- The assistant **never** sends orders or holds capital.
4. **Log (Trade tab)**
- After entering or exiting a trade, you log it in the assistant:
- Entry, exit, size, screenshots, rationale.
- This builds a complete dataset of your actual behavior.
5. **Review (Review tab)**
- Endofday/weekly review of:
- Equity curve and PnL.
- Patterns, strengths, weaknesses.
- Journal notes and habits.
- Use this to refine rules, checklists, and your playbook.
Every feature should reinforce some part of this loop.
---
## 5. Design principles
To keep the product aligned with the vision, we follow these principles:
1. **Assistant, not autopilot**
- AI can propose ideas, not promises.
- User makes the final decision.
- No direct execution or “oneclick trading.”
2. **Clarity over complexity**
- Fewer, more meaningful screens (Today / Trade / Review).
- Key information visible at a glance.
- Avoid overwhelming the trader during live market hours.
3. **Workflowfirst design**
- Every component should answer: “Where in the daily loop is this used?”
- Prefer features that enforce good process (checklists, journaling, review).
4. **Transparent AI**
- Show reasoning and context (not just signals).
- Make it clear what data the AI is using.
- Encourage the trader to **critically think**, not blindly follow.
5. **Safety and responsibility**
- Repeatedly remind: “This is decision support only.”
- Encourage proper risk management and capital allocation.
- No promises of profit, no “holy grail” marketing.
---
## 6. How docs and code should align with this vision
- **Docs**
- Must describe the app as an **assistant**.
- Must explicitly say that **trades are executed externally**.
- Any doc that suggests direct execution should be updated or archived.
- **Frontend**
- Tab names and labels should reflect this workflow:
- Today / Trade / Review.
- Buttons and labels: “Log Trade”, “Get AI Analysis”, “Plan”, “Review” — **not** “Execute Trade”.
- **Backend**
- Endpoints should be described as:
- `/ai/...`: analysis and plan generation.
- `/daily-helper/...`: structure, routines, checklists.
- `/analytics/...`: review, stats, patterns.
- No integration with broker APIs for order execution.
If a new feature or doc does not fit this vision, it should be redesigned or dropped.