Lucidly Pro — vault execution engine

Lucidly Provault execution engine

Platform

📖 Docs ↗

Active vault

Advanced

Live, read-only data from chain. Press ⌘K for the prompt console.

Vaults

All Lucidly Pro vaults

Lucidly Pro · no active vault

⌘

Create a Lucidly Pro vault

Add a custom chain

Lucidly Pro can deploy onto any EVM chain that has a LayerZero V2 endpoint. Look up the chain ID and EID at layerzero.network/deployments.

Chain name

Chain ID (decimal)

LayerZero V2 endpoint ID (EID)

Explorer URL

RPC URL

The chain is added in-memory to the wizard's selector — when you deploy, Lucidly's deployer service uses this RPC and the LayerZero teller reaches it via the EID.

Request a new strategy policy

Decomposer · model + API key

Pick a provider and paste the API key. Keys stay in your browser's localStorage — never sent to Lucidly. The decomposer runs entirely client-side; tool results (corpus / web search / contract reads) come from the live indexer + simulator + Etherscan v2 + Tavily APIs.

Model

API key for the selected provider

Optional research keys (improve novel-protocol decomposition)

Tavily key (web search)

Etherscan v2 unified key (verified ABIs)

Without these, the agent still works but can only use the corpus + your training-data fallback when a protocol isn't in the 4 reference strategies. Free tiers: tavily.com, etherscan.io/apis.

Anthropic in the browser: the SDK opts in via dangerousDirectBrowserAccess. Your key is sent directly to Anthropic; nothing routes through Lucidly. Use a key scoped to research workspaces if you want a separate audit trail.

Decomposing strategy

Trail simulator — preview against a chain fork. Never broadcasts.

🧠 Research

no provider configured

Provider tier

Groq API key

Get a free Groq key at console.groq.com/keys. The key is held in memory only; nothing is persisted.

Advanced settings

Indexer URL (Ponder GraphQL)

Active chain (for on-chain reads)

Model override (optional)

Ready 0 / 60,000 tokens