From bitkub-trade
Playbook for swing-trading the long side on Bitkub (THB pairs) using the gokub-mcp tools. Use when asked to find a trade, screen coins, evaluate an entry, size a position, place/manage an order, or review trading performance on Bitkub. Encodes the risk rules, the screen→regime→signal→size→validate→place→manage→log flow, and Bitkub-specific constraints (no native stop/OCO).
How this skill is triggered — by the user, by Claude, or both
Slash command
/bitkub-trade:playbookThe summary Claude sees in its skill listing — used to decide when to auto-load this skill
Long-only swing trading on Bitkub THB pairs via the gokub-mcp tools. The tools are
Long-only swing trading on Bitkub THB pairs via the gokub-mcp tools. The tools are the hands; this is the brain. Follow the flow, respect the guardrails, never skip the validation gate.
validate_trade_setup returns can_trade=true. This is the gate.client_side_stop_worker
(poll it on a loop until triggered=true). Never assume a resting stop exists.The analysis tools (check_market_regime, calculate_atr, calculate_rsi,
detect_breakout_signal, detect_pullback_signal, calculate_ema, calculate_roc)
take arrays (prices or OHLCV candles), not a symbol. So always:
get_historical_candles for the symbol+timeframe first.close prices (use extract_close_prices) or the candle array into the tool.Use the big timeframe (1D / 240) for regime + relative strength, the entry timeframe (60 / 15) for ATR + entry signals.
Before screening for new trades, check what you already hold.
get_trade_history(status_filter=open) — list all open positions.get_historical_candles(symbol, interval=60) → check_exit_signals(candles).should_exit=true or ≥ 2 exit reasons → exit immediately:
place_limit_order(side=sell) → log_trade_exit(trade_id, exit_price, exit_reason).check_trade_pnl(symbol, entry, qty, stop) — note current R to inform next steps.get_my_open_orders — collect all symbols that already have a pending limit order on
the exchange. Skip the entry step for any symbol already on this list. This prevents
duplicate orders when the skill runs on a recurring schedule (e.g. every 2H).Only after all urgent exits are handled, proceed to step 1 for new entries.
get_market_screener (volume, spread, depth) or get_market_overview
for context. Keep 5–10 liquid THB pairs.get_historical_candles (1D) → check_market_regime(prices).
Trade only if UPTREND. (Confirm with calculate_ema: close > EMA200, EMA50 > EMA200.)calculate_relative_strength_rank(symbols[], benchmark=btc_thb).
Keep the top names.calculate_atr(candles) → keep ATR% in a sane zone (~2–6%/day).detect_breakout_signal (close above 20-bar high
detect_pullback_signal (pullback to EMA20 + RSI bounce 40–50).get_wallet_balance → calculate_position_size(balance, risk_percent=2, entry, stop). Stop from signal's suggested_stop (swing low or 1.5×ATR).validate_trade_setup with the metrics above. Stop here if can_trade=false.get_symbol_rules → round_to_exchange_rules(symbol, price, qty). Reject
if valid=false (below min notional).simulate_trade to confirm R:R ≥ 2 before committing.place_limit_order(side=buy) then log_trade_entry → keep the
trade_id.client_side_stop_worker(symbol, side=sell, trigger=stop, limit_price, qty) and poll it (e.g. via /loop) until triggered.check_trade_pnl(symbol, entry, qty, stop) (fetches price; nets fees).client_side_stop_worker at entry).calculate_trailing_stop(candles, atr_multiplier=2); re-arm the worker at the new level.check_exit_signals(candles): STRUCTURE_BREAK or 2+ reasons →
exit now (place_limit_order(side=sell) or let the worker fire).log_trade_exit(trade_id, exit_price, exit_reason) (computes net PnL + R).get_trade_history(status_filter=closed) and calculate_expectancy. Expectancy =
win% × avgWinR − (1−win%) × avgLossR. If negative, change ONE variable and re-test.btc_thb; the library normalizes to THB_BTC for the API.place_limit_order amount: buy = THB to spend, sell = coin quantity.trades.json, override with env TRADES_FILE).Guides creation, editing, and verification of skills for AI coding agents using test-driven development with subagent scenarios. Use when authoring or debugging skills.
npx claudepluginhub dvgamerr-app/gokub-mcp --plugin bitkub-trade