Transparency

How Cladex actually trades

A plain-English breakdown of where an LLM sits in the system, how each strategy decides to buy and sell, and how you can verify every claim on this page for yourself. No black boxes.

The two layers of Cladex

Layer 1 — LLM (OpenAI)
UX only

A large language model reads your natural-language prompt (“I want a cautious BTC bot”) and turns it into a JSON config that picks one of our predefined strategies and one of four personalities.

The same model powers agent-to-user chat, so you can ask your agent what it’s doing and get a reply in character.

What it does not do: decide trades. The LLM has no path to the exchange. It cannot place orders, move funds, or change the strategy code.

Layer 2 — Trading engine (Python)
Deterministic rules

Five hand-coded strategies compute standard technical indicators (RSI, EMA, Bollinger Bands, ATR, volume ratios) and emit BUY/SELL signals when hardcoded threshold conditions are met.

Same market data in, same signal out. No hidden state, no model weights, no training loop. Every gate and threshold is visible in the strategy source files.

Every emitted signal ships with a reason string: exactly which conditions fired and what confidence score they produced.

The five strategies

Each strategy is a small, auditable Python file. Rules are listed in plain English here — full source in the repo.

DCA
Dollar-cost averaging on a schedule.
Personality · Any
Indicators used
Current price
Best for
Passive long-term holders who want automated accumulation without trying to time the market.
Entry
Buys a fixed dollar amount of each configured asset every cycle. No market-timing — just steady accumulation.
Exit
No automated exit. Positions build over time. Users close manually or via stop-loss set on the exchange.
Trend
Classic SMA-crossover trend following.
Personality · Any
Indicators used
Short SMA, Long SMA
Best for
Users who want a well-understood strategy with a small parameter surface.
Entry
BUY when the short moving average crosses above the long moving average. SELL when it crosses below. Standard trend-following logic.
Exit
Reverses on the opposite crossover. No pyramiding — one position per asset.
SafeFlow
Capital preservation with dip-buying in confirmed uptrends.
Personality · Nova
Indicators used
RSI(14), EMA(20/50/200), ATR(14) volatility filter
Best for
Conservative users. Priority is not losing money, not maximum return.
Entry
Requires: uptrend confirmation (EMA20 > EMA50 > EMA200), RSI in the neutral-to-oversold zone (not overbought), and volatility inside a filter band. Enters on the dip.
Exit
ATR-scaled stop-loss below entry. 3-tier take-profit ladder that scales out into strength.
TrendPro
Momentum trading with volume confirmation and breakout detection.
Personality · Sage
Indicators used
EMA crossovers, Volume ratio vs. 20-period average, Breakout detection
Best for
Balanced users who want more active engagement than SafeFlow without full aggression.
Entry
BUY when EMAs align in the trend direction AND current volume is above average AND price breaks the recent range with strength.
Exit
ATR-scaled stop-loss + 3-tier take-profit ladder. Scales out on the way up.
BeastMode
Explosive breakout hunter — high risk, high reward.
Personality · Apex
Indicators used
Bollinger Band squeeze, Volume accumulation trend, Breakout impulse strength, RSI(14), ATR(14)
Best for
Aggressive users who accept more losing trades in exchange for larger winners.
Entry
Fires only when THREE conditions align: price compression (BB squeeze, tight range, contracting ATR), volume accumulation (rising volume slope, above-average current volume), and a breakout candle with strong body and above-threshold impulse.
Exit
ATR-scaled stop-loss (2–8%). Take-profit ladder at +3% / +5% / +10% scaling out 33% / 33% / 34%.

The risk engine gate

Every signal a strategy produces passes through a separate risk engine before the trade reaches the exchange. The engine checks per-user exposure limits, monthly drawdown, cooldowns after consecutive losses, and account-level circuit breakers.

The behavior is fail-safe: if the risk engine is unreachable for any reason, the trade is blocked, not allowed through. The default is refusal to trade, never speculative execution.

After every executed trade, a post-trade check runs and can trigger cooldowns or risk-locks that will pause future entries.

How to verify all of this

1
Read the strategy source

Every strategy in worker/strategies/ is roughly 300 lines. All thresholds, gates, and confidence scoring are visible in the code. There is nothing else.

2
Read the per-signal reasons

Every trade in your history includes the reason string that produced it: which conditions fired, current RSI, volume vs. average, confidence score. No decisions are hidden.

3
Run paper mode

Paper mode is the default. It runs the full evaluate → risk-check → execute pipeline against simulated fills, so you can watch strategy behavior on your own account without a dollar at risk.

Questions this page didn’t answer?
We publish a live diagnostics endpoint and read-only strategy source. If something looks off, we want to hear about it.
Open diagnostics →