Session state
Knows whether the underlying exchange is open, closed, in pre-market, or approaching a scheduled event.
PRODUCT / RISK ENGINE
Nocturner evaluates the market around a proposed swap—not just the token price—then returns a score, a verdict, and the factors that produced it.
Session calendars, pool state, oracle updates, price relationships, filings, news, and halt feeds enter one pipeline.
Signals are normalized for the asset, market session, and proposed order size—then evaluated against policy.
The engine warns a human, answers an agent, publishes a feed, or blocks a transaction at the guard.
THE INPUTS
Knows whether the underlying exchange is open, closed, in pre-market, or approaching a scheduled event.
Measures how old the reference price is and escalates when the market is moving faster than the feed.
Compares your order to actual pool liquidity—not a headline TVL number that disappears on execution.
Tracks the distance between the token, oracle mid, and the underlying equity's last defensible close.
Reads filings, halt notices, corporate actions, and news that may not be reflected onchain yet.
THE OUTPUT
Humans and agents see the same verdict semantics. Policies decide what each verdict is allowed to do.
Conditions are inside policy limits.
Execution cost or divergence is elevated.
Reference data can no longer defend the price.
Policy rejects execution until conditions recover.
DELIVERY
Token watchlists, score history, factor explanations, and a clear trade-size-aware verdict.
Material catalyst alerts and a daily overnight gap report delivered where active traders already work.
Machine-readable queries and dry runs designed to sit directly inside an agent's planning loop.
Deterministic policy at the session-key level. A breached limit reverts—it is not merely logged.
A concise morning digest built from the same catalyst and risk pipeline.
Estimate price impact against indexed pool depth before signing anything.
Structured JSON and calendar feeds for splits, dividends, mergers, and ex-dates.
FAQ
No. AI discovers catalysts, extracts structured facts, and compiles intent. The final policy evaluation is deterministic and auditable.
No. It is a technical assessment of execution conditions and data quality, not a recommendation to buy or sell an asset.
Threshold calibration is the critical path. Historical observations and realized outcomes are needed before enforcement can be responsibly enabled.
No. The product is designed to work without a token. Any future token utility follows real guard usage rather than preceding it.
EARLY ACCESS / 2026
We're opening the read-only scorecard first while the engine collects the history needed for responsible calibration.