Retune Trading Bot Risk Caps in Real Time (Kiponos Python SDK)

Markets move faster than deploy pipelines. Your bot is live, volatility spikes, and you need to cut position size, tighten stop-loss, or halt new entries — now, not after the next release.

Kiponos.io gives Python trading systems live risk parameters: in-memory reads in the execution loop, WebSocket deltas from ops or a risk supervisor.

The execution-loop rule

def should_enter(kiponos, signal):
    risk = kiponos.path("risk", "intraday")
    if risk.get_bool("halt_new_entries"):
        return False
    if portfolio.drawdown() > risk.get_float("max_drawdown_pct"):
        return False
    return signal.strength > risk.get_float("min_signal_strength")

Every get_* is a local read — no HTTP between your bot and the exchange path.

Risk config tree

risk/
  intraday/
    max_position_usd: 25000
    max_drawdown_pct: 3.5
    stop_loss_pct: 1.2
    halt_new_entries: false
    min_signal_strength: 0.65
  symbols/
    BTC/
      max_position_usd: 10000
    ETH/
      max_position_usd: 8000

Supervisor + operator control

Actor	Action
Risk analyst	Lower `max_position_usd` in dashboard during news event
Supervisor algo	Set `halt_new_entries: true` when VaR exceeds limit
Trader	Per-symbol caps without restarting bot process