Background & Motivation

2.1 Evolution toward wallet-native agents

From bots to Autonomous Decentralized Finance

  1. Scripted bots (CeFi era). Single-purpose, exchange-specific scripts (REST/WebSocket) execute orders but lack shared state, composability, and formal safety rails.

  2. DeFi automation (keeper/MEV bots). On-chain triggers and arbitrage agents emerge, yet remain siloed per protocol/chain and are hard to govern.

  3. LLM tool-calling agents. Reason well but are “wallet-blind”: they propose actions humans must execute, re-introducing latency, custody risk, and operational toil.

  4. Wallet-native multi-agent systems (Degents). Cognition, coordination, and settlement live in one stack: agents reason, negotiate, and settle on-chain themselves under policy controls—laying the substrate for Autonomous Decentralized Finance and a true Agent OS for DeFi.

Limitations that persist without Degents

  • Fragmented SDKs for wallets, orchestration, and protocol calls

  • Human-in-the-loop custody and manual key ops

  • Ad-hoc compliance, auditing, and spending limits

  • Poor multi-agent coordination; strategy knowledge trapped in single bots

  • Slow iteration and unsafe rollouts due to lack of policy and simulation layers

2.2 Why wallet-native agents change the game

Definition — Autonomous Decentralized Finance (ADF). A market where fleets of software agents hold capital, form contracts, and settle obligations on public ledgers without centralized operators—governed by transparent policies and incentives.

Definition — Agent OS for DeFi. A runtime that makes ADF practical: secure key custody, chain-agnostic transactions, interoperable actions, multi-agent scheduling, stateful memory, and policy/compliance baked in.

Key shifts enabled by Degents

  • Trust-minimized settlement: Agents sign and broadcast on-chain; no off-chain handoffs.

  • Programmable policy: Spend limits, whitelists, KYC attestations, and circuit-breakers enforce guardrails at the runtime layer.

  • Composability at the action level: Any smart contract, CEX/DEX API, or AI skill becomes a first-class action callable via JSON or Python.

  • Always-on liquidity & credit: Agents rebalance, market-make, and route payments 24/7 with millisecond latency.

  • Economic agency: Agents can earn fees/yield, pay peers, and fund their own upgrades—bootstrapping self-sustaining services.

  • Observability & auditability: Every decision and signature is attributable, reproducible (via traces), and auditable on-chain.

Illustrative flow Signal → (multi-agent plan) → Risk/Policy check → Tx build/sign → On-chain execution → Post-trade learning → Strategy update.

2.3 Market gaps & opportunities

Gaps today

  • Custody & control: Either fully custodial (centralized risk) or DIY keys (operational risk); neither integrates with agent policy.

  • Orchestration debt: Teams stitch together LLMs, wallets, and protocol SDKs—fragile, hard to secure.

  • Compliance afterthought: KYC/AML, spend governance, and audit are bolted on, not intrinsic.

  • Limited collaboration: Bots don’t share beliefs, roles, or memory; cross-venue strategies degrade.

Opportunities unlocked by Degents

  • Agent-driven market making & liquidity orchestration across chains/venues with policy-constrained risk.

  • Autonomous treasuries for DAOs/enterprises: hedging, yield ladders, payroll, and supplier settlements.

  • Parametric finance (insurance, per-second subscriptions, streaming payments) with enforceable SLAs.

  • Cross-chain clearing & FX via intent-based routing and atomic settlement.

  • Creator & machine-to-machine economies: micropayments, revenue splits, and pay-per-compute between agents.

  • Compliance as code: attestations, sanctions screening, and audit trails embedded in the Agent OS.

Who benefits

  • Developers: a single stack to build, test, and ship ADF agents.

  • Funds/Desks: scale strategies by delegating to specialized agent swarms with built-in guardrails.

  • Enterprises/DAOs: verifiable policy enforcement and clean audit for on-chain finance.

  • Protocols: distribution via actions; instant reach to every Degents agent.

Thesis: By unifying cognition, coordination, and cryptographic settlement, Degents provides the missing Agent OS for DeFi, catalyzing the transition to Autonomous Decentralized Finance at internet scale.

The possibilities are endless with Degents. From Trading bots and prediction markets to personalized assistants, AI agents can be tailored to meet the needs of countless crypto use cases. Developers and creators are encouraged to innovate and define new applications, ensuring the ecosystem continues to grow and evolve.

Last updated