FarmDash Wagon Steward is a strictly read-only DeFi portfolio intelligence skill designed for OpenClaw autonomous agents. It serves as a non-custodial mirror of a user's on-chain farming state across multiple EVM networks, providing analytical depth without any transaction execution capabilities.
Core Functionality
The skill aggregates wallet balances through GET endpoints, returning structured data on total portfolio value, chain distribution, asset concentration, protocol positions, and idle capital. Its flagship get_capital_efficiency tool produces a 0-100 score with transparent formula breakdown, penalizing idle stablecoins, over-concentration above 40%, stale positions beyond 90 days, and exposure to cooled protocols while rewarding yield-generating allocations. The get_rebalance_plan tool generates swap proposals against user-defined target allocations, though execution remains strictly outside this skill's scope.
Significant Advantages
Zero custody architecture eliminates private key exposure entirely—Wagon Steward only accepts public wallet addresses and optional Bearer tokens for rate-limit tiers. The explicit read-only posture (every tool is GET-only) creates a hardened security boundary where even complete compromise of the MCP server could not drain user funds. Multi-chain aggregation (Ethereum, Base, Arbitrum, others via Alchemy) provides unified portfolio visibility rare in fragmented DeFi tooling. The transparent scoring heuristic allows users to audit and override recommendations with full visibility into calculation logic. Composition design enables clean hand-offs to execution skills (Signal Architect for swaps, Futures Strategist for hedges) under separate ClawScan-reviewed contracts.
Limitations & Constraints
Wallet portfolio endpoints require paid Pioneer tier ($29/month); free Scout tier offers only 5 requests/24h with public discovery endpoints only, making meaningful portfolio analysis paywalled. Data staleness is inherent—on-chain reads are point-in-time snapshots vulnerable to mempool shifts within minutes. Capital efficiency scoring remains a heuristic, not financial advice; the formula's weightings (30% idle penalty, 20% concentration penalty) may not match individual risk preferences. Unverified tokens in cross-chain holdings may go unpriced, creating blind spots. Rate limits (500/day Pioneer, 50k/day Syndicate) throttle high-frequency monitoring use cases.
Ideal Users
DeFi active farmers managing multi-chain positions who want analytical coherence without surrendering execution control; OpenClaw agent operators building autonomous treasury management workflows; users comfortable with subscription pricing for portfolio intelligence; those prioritizing non-custodial verification over custodial convenience.
Risk Considerations
The skill cannot prevent user error in downstream execution—the separately installed execution skills carry their own smart contract risks under ClawScan review, but Wagon Steward's rebalancing proposals assume ideal liquidity conditions that real markets may not provide. Opportunity cost calculations are illustrative only; realized yields depend on execution-time conditions. The "read-only" branding, while accurate for this skill, may create complacency about risks in companion execution tools. Agent misconfiguration could lead to stale data driving decisions—explicit asOf timestamps and staleAfterMs hints require orchestrator respect. Bearer tokens, while tier-only, do create an authentication surface that should be rotated if suspected compromised.