About Wickra Copilot

Wickra Copilot is a local market copilot grounded in real microstructure. It folds live order-book, liquidation and funding data into a MarketContext — a list of hard, numeric facts — and hands that grounding to the LLM of your choice, so an answer to "why did BTC just dump?" is anchored to the real book, not vibes.

What makes it different

• Grounded, not generated. A MarketContext is a pure function of the feed: price moves, order-book imbalance, liquidation clusters, funding flips, OI changes and volatility spikes, each with a ready-made human sentence.
• The core is deterministic and offline. Fact derivation has no network, no key and no I/O, and is byte-identical across all ten bindings and both build profiles. The network call lives in a separate crate.
• LLM-agnostic, your own key. Ollama runs fully offline; OpenAI, Claude or Gemini use your own key over their endpoints — or skip the LLM and read the facts directly. Not a hosted service, not a SaaS.
• Read-only. It reads market data and asks questions; it never places orders.

Why it exists

LLMs hallucinate market state. Wickra Copilot fixes the grounding: it computes the facts first, deterministically, then lets a model narrate them. It defines the context surface once, in Rust, and exposes it as a JSON-over-C-ABI data API to Rust, Python, Node.js, WASM and — over a C ABI — C, C++, C#, Go, Java and R.

Open source

Released under the MIT OR Apache-2.0 license — permissive, OSI-approved, free for any use including commercial. Source, issues and releases on GitHub.

Disclaimer

Wickra Copilot is a software library, not a trading system, and is provided as-is with no warranty. It surfaces facts and narrates them; it does not give financial advice. Use it at your own risk.