Ingest EVM market data
Stage 1 focuses on allowlisted EVM pool state, prices, liquidity, flows, and token-risk inputs.
Keele Labs is building an EVM-first AI market-analysis system for allowlisted pairs. Stage 1 focuses on reliable data and evidence before signals, user-directed execution, or DEX integrations are considered.
EARLY-STAGE PRODUCT / FEATURES, INTEGRATIONS, AND AVAILABILITY MAY CHANGE
MARKET
ALLOWLISTED EVM PAIR
CURRENT SCOPE
MARKET ANALYSIS
Market data
EVM pool, price, and flow inputs
Feature engineering
Depth, liquidity, and regime context
Evidence-bound analysis
Structured AI reasoning layer
Signals & execution
Not enabled during Stage 1
DATA QUALITY
EVIDENCE-BOUND
EVM-FIRST
ILLUSTRATIVE STAGE 1 INTERFACE / NOT LIVE MARKET DATA OR EXECUTION
How it works
Stage 1 establishes the data and analysis foundation. Any later model proposal remains separate from the deterministic policy that decides what may proceed.
Stage 1 focuses on allowlisted EVM pool state, prices, liquidity, flows, and token-risk inputs.
The planned feature layer covers volatility, trend, depth, slippage, liquidity health, flow, and market regimes.
AI analysis is being designed around compact, validated market evidence rather than raw candles or price predictions.
Provider health, raw snapshot retention, bias guards, and reproducible analysis are required before signals are considered.
Stage 1 capabilities
Build a structured view of allowlisted EVM market conditions before generating trading signals.
Require structured inputs, explicit evidence, and a deterministic fallback for planned AI analysis.
Later execution stages are planned to enforce asset, venue, position, exposure, and slippage boundaries.
Target architecture keeps assets in user-controlled wallets with scoped, revocable permissions.
DEX integrations remain under technical, liquidity, and security review; none are announced as supported.
Monitoring of signals, transactions, positions, and risk events is planned for later execution stages.
Trust & security
Keele Labs is being designed around scoped wallet permissions, explicit trading boundaries, and records of the policy checks and transaction events associated with each action.
Read the security modelThe target architecture keeps assets in user-controlled wallets rather than pooled accounts.
Later permissions and limits are intended to define what the agent may attempt. They do not guarantee a profitable outcome.
When transactions are submitted in a future execution stage, they will remain publicly verifiable. Transparency does not remove protocol or market risk.
Keele Labs cannot eliminate market, liquidity, smart-contract, bridge, oracle, network, execution, model, or key-management risk. Transactions may fail, be reordered, or execute at worse prices than expected. You can lose some or all assets traded.
Production operations
Keele Labs is still an early-stage product. Before user access, the target system needs health checks, automatic pauses, replayable records, and documented recovery procedures.
01 / SYSTEM HEALTH
Track data freshness, RPC response, signer availability, simulation results, and transaction confirmation state.
02 / SAFE PAUSE
The target operating model pauses execution when a dependency, policy check, or exposure limit reports an unsafe state.
03 / TRACE & RECOVER
Replayable decision and transaction events should make incidents reviewable before execution is resumed.
Production readiness will depend on implemented architecture, security review, operational testing, and incident procedures. This block describes intended standards, not current availability.
FAQ
Early access
Join the early-access list for product updates and testing invitations. Joining does not guarantee access.
NO WALLET ADDRESS REQUIRED / FORM STORAGE NOT CONNECTED