How meet252501/yeno-volume-bot-challenge- works
YeNo Volume Bot is a simulated trading agent for BTC five-minute Up/Down markets. Its main workflow accepts market, account, reference-price, and order-book state, then returns HOLD, BUY, or SELL; the strategy must manage fees, risk, completed-cycle volume, and finishing flat. The repository has a modular reference /decide implementation and a separate, more elaborate Velocity bot in bot.py. Replay reporting is present, but replay execution is explicitly unimplemented. The architecture below reflects sampled code and README capabilities; some documented strategy subsystems have no sampled implementation wiring.
Main entry points: server.py Reference API · decision.py Decision engine · bot.py Velocity HTTP handler · replay.py Replay harness
Components 14
Decision interfaces
- Decision request
src/models.py - Reference API
src/api/server.py - Decision response
src/models.py - Velocity HTTP handler
bot.py - Health status
src/api/server.py
Trading strategy
- Velocity bot
bot.py - Decision engine
src/decision.py - Fee model
src/fees.py
Risk and accounting
- Risk controls
src/risk.py - Cycle tracker
src/cycle_tracker.py
Replay evaluation
- Replay paths
data/replay - Replay harness
src/backtest/replay.py - Replay report
src/backtest/report.py
Other components
- Challenge evaluator
Connections 13
- Challenge evaluator → Reference API: calls /decide
- Reference API → Decision request: validates as
- Reference API → Decision engine: calls
- Reference API → Cycle tracker: passes state
- Decision engine → Risk controls: checks sizing
- Decision engine → Fee model: scores price
- Decision engine → Cycle tracker: checks flatness
- Decision engine → Decision response: returns action
- Reference API → Health status: reports volume
- Challenge evaluator → Velocity HTTP handler: posts observation
- Velocity HTTP handler → Velocity bot: calls decide
- Replay paths → Replay harness: loads paths
- Replay harness → Replay report: builds report