WhyBot — Why did the bot do that?

touch_app

Pick a coin from the list

Or run a live check first to pull prices, order book, and trade flow from Bitget.

Overview & trends Total checks: — · Bitget pairs: — expand_more

What the bot picked

Safety grades

Safety score over time

Higher = safer decision. Green/red dots = buy/sell checks.

Past checks

Search and filter older decisions · j/k to move

Problems only

Connect your own agent

WhyBot does not run your bot — it records and grades each decision. POST one JSON object after every trade (or skip/hold).

curl -X POST https://whybotai.vercel.app/api/decisions \
  -H "Content-Type: application/json" \
  -d '{
    "agentId": "my-bitget-bot",
    "agentName": "My Strategy Bot",
    "thesis": "Flat market — wait",
    "reasoning": "24h move small, RSI neutral.",
    "confidence": 70,
    "inputs": { "symbol": "BTCUSDT", "price": 63000 },
    "action": { "type": "HOLD", "symbol": "BTCUSDT" },
    "tags": ["LIVE"]
  }'

Node/TS agents can use AutopsySession from src/core/logger.ts. See README → Connect your agent.

Check market runs the built-in demo bot only (Live Regime Router). External agents use POST /api/decisions.

What do these words mean?

Safety score: 0–100 report card. Higher = safer bot choice.
HOLD / BUY / SKIP: What the bot wanted. Paper only — no real trades.
Momentum (RSI): Our math from Bitget price history. Hot or cold market.
Order book / flow / OI: Live public Bitget data. Re-check a coin to refresh.
Proof log: JSON of real API calls for hackathon judges.
Download PNG: Shareable card image of one decision.