Wickra CopilotA local market copilot.

The context is JSON, not code ​

A ContextSpec names the symbols to inspect, a lookback window and the facts to derive. The builder walks each symbol's feed and returns a stable, self-explaining MarketContext.

{
  "symbols": ["BTCUSDT"],
  "lookback": 20,
  "timeframe": "1m",
  "facts": ["price_move", "orderbook_imbalance", "liquidation_cluster", "funding_flip", "oi_change", "volatility_spike"]
}

Each Fact carries a signed value, a ranking magnitude, a timestamp and a ready-made human sentence (e.g. "BTCUSDT dropped -6.44% over the last 20 bars."), so the context explains itself before any LLM sees it.

Install ​

The same copilot from every language — native Rust, Python, Node.js and WASM, plus a C ABI for C, C++, C#, Go, Java and R.

pip install wickra-copilot

Grounding first, generation second ​

Build the deterministic MarketContext from the feed, then hand it to the LLM of your choice — or none. The command API returns the same bytes in every binding.

import json
from wickra_copilot import Copilot

spec = json.dumps({
    "symbols": ["BTCUSDT"],
    "lookback": 20,
    "facts": ["price_move", "liquidation_cluster", "funding_flip"],
})

copilot = Copilot(spec)
context = json.loads(copilot.command(json.dumps({"cmd": "build_context", "feeds": feeds})))
for fact in context["facts"]:
    print(fact["human"])

Built on the Wickra core ​

Wickra Copilot is part of the Wickra ecosystem. Its facts are derived from the same typed microstructure feeds that wickra-core and wickra-exchange produce, so the copilot reasons over exactly the numbers a backtest or a live chart would see.

Wickra Copilot is a software library, not a trading system, and gives no financial advice — use at your own risk.