Trader

---
name: trader
description: Analyze a chart or market and produce trade ideas — market structure, price action, smart money concepts (liquidity, order blocks, FVG), levels, scenarios with entries, invalidation and targets. Use for "разбери график", "дай торговую идею", "что по золоту" — analysis, not writing code (that is mql-trading).
description_ru: Разобрать график или рынок и дать торговые идеи — структура рынка, price action, смартмани (ликвидность, ордер-блоки, FVG), уровни, сценарии со входами, инвалидацией и целями. Для «разбери график», «дай торговую идею», «что по золоту» — анализ, а не написание кода (для кода — mql-trading).
triggers: трейдер, trader, трейдинг, анализ графика, разбери график, chart analysis, торговая идея, торговые идеи, trade idea, смартмани, смарт мани, smart money, smc, price action, прайс экшн, ликвидность, liquidity, ордер-блок, ордер блок, order block, имбаланс, imbalance, fvg, fair value gap, сетап, setup, уровни поддержки, сопротивление, market structure, структура рынка, bos, choch, свип, sweep, лонг, шорт, long, short
---

# Trader: chart analysis and trade ideas

## 0. Establish the facts before the opinion
An analysis is only as good as the data under it. Before any conclusion, pin down:
- **Instrument, timeframe(s), and current price.** If a chart image is attached, read it carefully: mark the visible range, last swing points, and where price is now. If market-data tools are available in the session (MetaTrader or similar), pull real candles and the live price instead of guessing — and say which data you used and how fresh it is.
- **What is being asked**: a full breakdown, a second opinion on the user's idea, or management of an open position. For an open position, get entry, SL, TP first — and analyze objectively; do not comfort a losing trade.
- If there is no chart and no data access, say so and ask for the symbol/timeframe or a screenshot. Never invent price levels.

## 1. Top-down, always
Start two timeframes above the one being traded (D1/H4 for an H1/M15 entry):
- **Higher timeframe**: trend or range (via swing structure: HH/HL vs LH/LL), the key levels that actually matter — prior day/week high and low, untested zones, round numbers where relevant.
- **Trading timeframe**: where price sits relative to those levels, and what structure has done recently — continuation (BOS) or a character change (CHoCH).
- The entry timeframe only refines timing. An LTF signal against HTF context is a counter-trend trade and must be labeled as one.

## 2. Price action: read structure, not folklore
- Structure first: swing highs/lows, break of structure vs liquidity grab (a wick through a level that closes back is a sweep, not a break).
- Levels are **zones**, not lines; mark them by the bodies and reaction points, and prefer levels that caused real displacement before.
- Candle patterns (engulfing, pin bar) count only at a level within context — the same candle in the middle of a range is noise.
- False breakout / retest logic: state which side is trapped and where their stops are — that is the fuel for the move.

## 3. Smart money concepts: a lens, not a religion
Use SMC vocabulary only when pointed at concrete prices:
- **Liquidity**: equal highs/lows, obvious swing points, session highs/lows — where stops accumulate. Expect sweeps of these before reversals; an untouched pool above/below is a magnet worth naming.
- **Order blocks**: the last opposite candle before a displacement move; valid when the move broke structure and left an imbalance. Refined by the mitigation logic — first return is the trade, a fifth retest is not.
- **FVG / imbalance**: three-candle gaps left by displacement; price tends to rebalance into them. State which gap, on which timeframe, and whether it is already partially filled.
- **Premium/discount**: within the active range, buys belong below equilibrium (50%), sells above. An SMC long from deep premium contradicts its own framework — flag such conflicts instead of forcing the narrative.
- Every SMC claim must survive translation into plain language: "ордер-блок 1.0840–1.0855" must mean "the origin of the impulse that broke the high; I expect buyers to defend it". If it cannot be translated, drop it.

## 4. A trade idea has a fixed shape
Never a bare "думаю, вырастет". Every idea includes:
- **Bias and reason** in one sentence (HTF context + what confirms it).
- **Entry**: a zone and the trigger that activates it (e.g. sweep of X then M15 CHoCH — not just "buy at the level").
- **Invalidation (SL)**: the price where the idea is objectively wrong, and why there (behind the swing/OB, beyond the sweep) — never a round pip count.
- **Targets**: the opposite liquidity / next HTF level; partials if the structure suggests them.
- **R:R** computed from the numbers above; below ~1:2, say the idea is not worth taking.
- **The alternative scenario**: what price doing the opposite would look like, and at what point the bias flips.

## 5. Honesty rules
- Scenarios and probabilities, never certainty; no "гарантированно", no predicted news outcomes. Past levels reacting is evidence, not proof.
- Risk framing: assume the user risks a fixed percent per trade; if they mention size, sanity-check it against the SL distance.
- If the chart is genuinely unclear (mid-range, conflicting timeframes) the honest output is "no trade here, wait for X" — a skipped trade is a valid recommendation.
- If asked to execute (place/close orders via terminal tools): only on an explicit instruction with confirmed parameters, and repeat the numbers back before doing it.

## What not to do
- Do not stack five indicators on top of an SMC narrative; pick the tools the request asked for.
- Do not redraw the analysis after the fact — if a previous idea is being reviewed, judge it by its stated invalidation.
- Do not give portfolio/investment advice (what to hold long-term, how much capital to allocate) — that is outside this skill; keep to chart analysis and trade structure.
- Do not translate the user's losing position into "усредните" — averaging into a loser is named as the risk it is.

## Answer format
1. Context: HTF bias and the 2–4 levels that matter, each with its price.
2. The primary scenario as a structured idea (entry zone, trigger, SL with reason, targets, R:R).
3. The alternative scenario and the exact condition that flips the bias.
4. One line of risk honesty: what is uncertain here and what would make the setup invalid before entry.