Authors: 0xjacobzhao and ChatGPT 4o
In the current crypto industry, stablecoin payments and DeFi applications are among the few tracks that have been verified to have real demand and long-term value. At the same time, the blooming agents are gradually becoming the actual landing form of the AI industry aimed at user interfaces, becoming the key intermediary layer connecting AI capabilities and user needs.
In the fusion field of Crypto and AI, especially in the direction where AI technology feeds back into Crypto applications, current explorations are mainly focused on three typical scenarios:
Conversational interaction-type agents: Primarily chat, companionship, and assistant types. Although most are still general large models, the low development threshold and natural interaction, combined with token incentives, have made them the earliest forms pushed to the market for user attention.
Information integration-type agents: Focus on the intelligent integration of online and on-chain information. Kaito, AIXBT, etc., have succeeded in online but non-chain information search integration fields, while on-chain data integration direction is still in the exploratory stage with no obvious standout projects.
Strategy execution-type agents: Extending Agent Payment and DeFAI in depth, with a focus on stablecoin payments and DeFi strategy execution. Such agents are more deeply embedded in on-chain trading and asset management logic, expected to break through speculation bottlenecks and form intelligent execution infrastructures with financial efficiency and sustainable yields.
This article will focus on the evolution path of the fusion between DeFi and AI, outlining the development stages from automation to intelligence, and analyzing the infrastructure, scenario space, and key challenges of strategy execution agents.
Three stages of DeFi intelligence: Automation, Copilot, and the leap to AgentFi
In the evolution of DeFi intelligence, we can divide system capabilities into three stages: Automation (Automation Tools), Intent-Centric Copilot (Intent-Centric Assistant), and AgentFi (On-chain Intelligent Agents).
Automation is more like a rule trigger: executing fixed tasks based on preset conditions, such as arbitrage, rebalancing, stop-loss, etc., cannot generate strategies, nor can it operate independently.
Copilot introduced intent recognition and semantic parsing capabilities, allowing users to input natural language, with the system understanding, breaking down, and suggesting execution paths, but ultimately still requiring user confirmation, and the execution chain is not closed.
AgentFi represents a complete 'perception → reasoning/strategy generation → on-chain execution → evolution' intelligent closed loop, being an intelligent agent with on-chain autonomous execution and continuous evolution capabilities.
Dimension Automation Infra Intent-Centric Copilot AgentFi Core Logic Rule triggered + conditional execution Intent recognition + operation guidance Strategy closed loop + autonomous execution Execution Method Triggered execution based on preset conditions (if-then) Understand user instructions, assist in breaking down operations Fully autonomous perception, judgment, execution User Interaction No interaction required, execution passively triggered Users express intent through prompts, system assists in breaking down No human interaction required, can collaborate with humans/agents Intelligent Level Low, process automation Medium, capable of interactive understanding High, autonomous strategy generation and evolution Strategy Capability None, executes preset tasks Limited, relies on user instructions Strong, capable of self-learning and optimizing combinations Implementation Difficulty Low, backend service-oriented Medium, strong frontend interaction design required High, deep collaboration required between AI/execution infrastructure On-chain Execution ✅ Perception
❌ Decision (fixed rule triggering)
✅ Supports simple execution
✅ Perception
✅ Decision
⚠️ Execution requires user-assisted confirmation.
✅ Perception
✅ Decision
✅ Complete closed-loop on-chain execution
Typical representatives include Gelato, Mimic, HeyElsa.ai, Bankr, Giza ARMA.
To determine whether a project truly belongs to AgentFi, one must see if it meets at least three of the following five core standards:
Autonomous perception of on-chain states/market signals (not static inputs, but real-time monitoring)
Possesses strategy generation and combination capabilities (not preset strategies, but can self-develop action plans based on context)
Can autonomously perform on-chain operations (no user interaction required, able to execute complex operations like swap/lend/stake)
Possesses persistent state and evolutionary capabilities (agents have a lifecycle, can operate long-term and self-adjust based on feedback)
Possesses an Agent-Native architecture (such as dedicated Agent SDK, custodial execution environment, Agent middleware, etc.)
In other words, automated trading ≠ Copilot, and even less ≠ AgentFi: Automated trading is merely a 'rule trigger', while Copilot can understand user intent and provide operation suggestions but still relies on human participation; true AgentFi is 'an intelligent agent with perception, reasoning, and on-chain autonomous execution capabilities', capable of completing strategy closed loops and continuous evolution without human intervention.
DeFi scene intelligent adaptation analysis:
In the DeFi (Decentralized Finance) system, core application scenarios can roughly be divided into asset circulation and exchange types and yield-type finance types. We believe that the adaptability of these two categories of scenarios on the intelligentization path shows significant differences:
1. Asset circulation and exchange scenarios
Asset circulation and exchange scenarios primarily involve atomic interactions, including Swap trading, cross-chain bridges, fiat entry and exit, etc. Their essential characteristic is 'intent-driven + single atomic interaction', where the trading process does not involve yield strategies, state maintenance, or evolutionary logic, mostly suitable for the lightweight execution path of Intent-Centric Copilot and do not belong to AgentFi.
Due to its low engineering threshold and simple interaction, the majority of DeFAI-type projects on the market are currently at this stage, which do not constitute closed-loop intelligent agents of AgentFi; however, a few advanced complex swap strategies (such as cross-asset arbitrage, perpetual hedging LP, leveraged rebalancing, etc.) indeed require AI agent capabilities to integrate, currently still in the early exploratory stage.
Scenario Category Is there continuous yield? AgentFi Adaptability Engineering Implementation Difficulty Description Swap Trading ❌ No ⚠️ Partial adaptation (only Intent trading is not true AgentFi) ✅ Easy to implement Single atomic operations (such as token swaps), no strategy state accumulation, suitable for Copilot invocation. Cross-chain bridge ❌ No ❌ Weak ✅ Easy to implement Cross-chain is intermediary transmission, does not involve strategy planning and adjustment, AI participation is marginal. Fiat entry and exit ❌ No ❌ None ❌ Uncontrollable Highly dependent on CeFi channels and compliance processes, on-chain agents cannot autonomously initiate operations Aggregation optimization ⚠️ Not necessarily ⚠️ Partial adaptation ✅ Medium Mainly based on automation tools, if able to combine multi-platform quotes or maximize yield paths, could be executed by lightweight agents, but difficult to evolve into intelligent agents ✅ Swap trading combinations ✅ Possible yields ✅ Not mature ❌ Difficult to implement Such as cross-asset arbitrage, perpetual hedging LP, dynamic position allocation, etc., require complex strategy engines to support, currently still in prototype stage with no available agents.
2. Asset yield-type financial scenarios
Financial scenarios focused on asset yield have clear yield objectives, complex strategy combination spaces, and dynamic state management needs, which naturally fit the 'strategy closed loop + autonomous execution' model of AgentFi. Its core features are as follows:
Quantifiable yield objectives (APR / APY) make it easy for agents to establish optimization functions;
Strategy combination space is vast, covering multiple assets, multiple maturities, multiple platforms, and multiple interaction processes;
Operations require frequent management and real-time adjustments, suitable for execution and maintenance by on-chain intelligent agents (Agents).
Ranking Scenario Category Continuous Yield? AgentFi Adaptability Engineering Difficulty Description 1 Yield Farming ✅ Yes ✅✅✅ Extremely High ❌ High Strategies need frequent dynamic adjustments (such as reinvestment, migration, dual-pool strategies, etc.), most suitable for deploying AI strategy agents 2 Lending ✅ Yes ✅✅✅ Extremely High ✅ Low Interest rate fluctuations + collateral status can be read, risk warnings and automatic portfolio adjustments are easy to achieve 3 Pendle (PT/YT yield rights trading) ✅ Yes ✅✅ High ❌ High Yield periods and structures are diverse, combination trading is complex, agents can optimize buy/sell timing and yield stability 4 Funding rate arbitrage (Perp/CeFi/DeFi mixed) ✅ Yes ✅✅ High ❌ Extremely High Multi-market arbitrage has AI advantages, but off-chain interaction and collaboration complexity is extremely high, still in the exploratory stage 5 Staking / Restaking / LRT strategy combination ⚠️ Fixed yield ⚠️ Conditionally adaptable ⚠️ Medium Static staking is not suitable for agents, but multiple LST + Lending + LP dynamic combinations for agents may intervene 6 RWA (real-world assets) ⚠️ Stable yield ❌ Low ⚠️ Compliance heavy Yield structures are stable, compliance thresholds are high, protocols do not interconnect, short-term AgentFi strategy landing space is not available.
Due to multiple factors such as yield duration, volatility frequency, on-chain data complexity, cross-protocol integration difficulties, and compliance restrictions, the adaptability and engineering feasibility of different yield scenarios in the AgentFi dimension show significant differences, with the following priority recommendations:
High-priority business landing direction:
Lending (Lending/Borrowing): Interest rate fluctuations are easy to track standardized execution logic, suitable for lightweight intelligent agents.
Yield Farming: Pools are dynamically frequent, with a large strategy combination space and high yield fluctuations; AgentFi can significantly optimize annual returns and interaction efficiency, but engineering implementation has certain challenges.
Mid-to-long-term exploratory layout directions:
Pendle yield rights trading: The time dimension and yield curve are clear, suitable for agents to manage expiration rotations and pool arbitrage;
Funding rate arbitrage: Theoretical yields are considerable, but the cross-market execution and off-chain interaction challenges must be addressed, with high engineering difficulty;
LRT dynamic combination structure: static staking is not suitable, but attempts to automatically adjust LRT + LP + Lending strategies may exist.
RWA multi-asset combination management: Difficult to land in the short term, agents can provide assistance in combination optimization and expiration strategies.
Project introduction for intelligentization in DeFi scenarios:
1. Automation Tools (Automation Infra): Rule-triggered and condition-executed
Gelato is one of the earliest infrastructures for DeFi automation, having provided conditional trigger task execution support for protocols like Aave and Reflexer, but it has now transformed into a Rollup as a Service provider. Currently, the main battleground for on-chain automation has also shifted to DeFi asset management platforms (DeFi Saver, Instadapp). These platforms integrate standardized automatic execution modules, including Limit Order setting, liquidation protection, automatic rebalancing, DCA, grid strategies, etc. Additionally, we see some more complex DeFi automation tool platform projects:
Mimic.fi (https://www.mimic.fi/)
Mimic.fi is an on-chain automation platform serving DeFi developers and projects, supporting the construction of programmable automated tasks on chains like Arbitrum, Base, and Optimism. Its core achieves cross-protocol automated execution through 'if-then' rule triggers, with an architecture divided into Planning (task and trigger definition), Execution (intent broadcasting and execution bidding), and Security (triple verification and security control) layers. Currently adopting an SDK access method, the product is still in the early deployment phase.
AFI Protocol（https://www.afiprotocol.ai/）
AFI Protocol is an algorithm-driven Agent execution network that supports 7×24 hours of non-custodial automated operations, focusing on solving issues of execution dispersion, strategy thresholds, and risk responses in DeFi. Its design targets institutions and advanced users, providing orchestrable strategies, permission management, and SDK tools, and introduces the yield-generating stablecoin afiUSD as its native asset. Currently in the Sonic Labs internal testing phase, it has yet to publicly launch or open for retail user access.
2. Intent-Centric Copilot: Intent Expression and Execution Suggestions
The once hot DeFAI concept at the end of 2024, excluding some speculation primarily involving meme tokens, essentially most projects belong to the Intent-Centric Copilot type—expressing user intent through natural language, with the system providing trading suggestions or completing basic on-chain operations. Its core capabilities still remain at the 'intent recognition + Copilot-style auxiliary execution' stage, and have not formed a complete strategy closed loop or continuous optimization mechanism. Many products have notable shortcomings in semantic understanding, cross-protocol invocation, and feedback response, with generally poor user experience and relatively limited functional boundaries.
HeyElsa (https://app.heyelsa.ai/)
HeyElsa is an AI Copilot positioned for Web3 scenarios, empowering users to complete various on-chain operations including trading, cross-chain bridging, NFT purchases, stop-loss settings, and Zora token creation through natural language interaction. As a multifunctional conversational crypto assistant, it covers users from beginners to advanced traders (including highly active degen groups), currently supporting real-time interactions on over 10 mainstream blockchains. The platform's daily trading volume has reached $1 million, with daily active users between 3,000 and 5,000, and the system has integrated yield optimization strategies and automated intent execution modules, initially building the foundational capability framework for AgentFi applications.
Bankr (https://bankr.bot/)
Bankr is an intent trading assistant integrating AI, DeFi, and social scenarios, allowing users to issue commands in natural language via the X platform or dedicated terminals to complete operations like Swap, limit orders, cross-chain bridging, token issuance, and NFT minting, supporting Base, Solana, Polygon, and Ethereum mainnet. Bankr has built a complete Intent → Compile → Execute chain, emphasizing a minimalist trading experience and seamless operation within social environments, activating the ecosystem through token incentives and profit-sharing mechanisms.
Griffain (https://griffain.com/)
Griffain is a multifunctional AI agent platform deployed on Solana, supporting user interactions with Griffain Copilot through natural language to execute on-chain operations such as asset queries, swaps, NFT trading, and LP management. The platform has multiple intelligent agent modules built-in and encourages community participation in agent creation and sharing. Technically built on the Anchor Framework and components like Jupiter and Tensor, it emphasizes mobile adaptation and frontend composability. Currently, it supports over 10 core agent modules, with strong execution capabilities and ecological linkage.
Symphony (https://www.symphony.io/)
Symphony is an on-chain execution infrastructure for AI agents, building a full-stack system covering intent modeling, intelligent path discovery, RFQ execution, and account abstraction, aiming to become the core module of the DeFi intelligent execution layer. The platform has launched the conversational assistant Sympson, which has market query and strategy suggestion functions, but on-chain execution has not yet been opened. Symphony provides the core components needed for AgentFi, which can support multi-agent collaborative execution and cross-chain operations in the future.
Hey Anon (https://heyanon.ai/)
HeyAnon is a DeFAI platform that combines intent interaction, on-chain execution, and intelligence analysis, supporting multi-chain deployment (Ethereum, Base, Solana, etc.) and cross-chain bridging (LayerZero, deBridge). Users can perform operations such as Swap, lending, and Staking through natural language, and obtain on-chain sentiment and market dynamics analysis. Despite the project's high attention due to founder Sesta, it is still in the Copilot stage, and the core strategy and execution intelligence have not yet fully landed, requiring further observation for long-term development.
Project Name Core Positioning On-chain Execution Feasibility Review Rating HeyElsa Conversational DeFi assistant for beginners or degens ⚠️ Limited (basic on-chain interaction) Intent→Agent→Execution closed loop has initially formed, with the strongest multi-chain support capability, friendly interaction, and clear positioning. 4 Bankr Natural language trading assistant + social embedding ⚠️ Limited (Beta 0.5) Can perform basic trading operations, interface is crude but closed loop is taking shape, user social embedding experience is prominent 3.5 Griffain Solana multi-agent Copilot platform ⚠️ Limited (basic functions) Supports multi-module invocation, but strategy combination and intelligence are weak, Solana restricts its multi-chain adaptability 3.0 Symphony Cross-chain execution infrastructure + account abstraction ❌ None (only conversational suggestions) Solid architecture, contains key modules required for AgentFi execution, but has not interfaced with actual user scenarios 2.0 HeyAnon Multi-chain DeFi intent assistant + market intelligence analysis ❌ None (only text dialogue) The product remains at the Q&A level, with no real on-chain capabilities, has not formed an execution closed loop, and contains a certain market hype component 1.5
The above rating system is mainly based on the current usability of the product, user experience, and the feasibility of executing the publicly available roadmap, which has a certain subjectivity. Please note that this assessment does not involve code security checks and does not constitute investment advice; thank you for your understanding.
3. AgentFi agents: Strategy closed loops and autonomous execution
We believe that AgentFi represents a more advanced form of DeFi intelligence leap compared to Intent Copilot. Agents possess independent yield strategies and on-chain autonomous execution capabilities, significantly enhancing users' strategy execution efficiency and capital utilization rates. By 2025, we are excited to see more and more AgentFi projects that have landed or are planning products, mainly focusing on lending and yield farming directions, with representative projects including Giza ARMA, Theoriq AlphaSwarm, Almanak, Brahma, Olas series, etc.
Giza ARMA(https://arma.xyz/)
ARMA is an intelligent agent product launched by Giza, designed for optimizing cross-protocol yields for stablecoins. It is deployed on the Base network, supporting multiple mainstream lending protocols such as Aave, Morpho, Compound, and Moonwell, with core capabilities including cross-protocol rebalancing, automatic compounding, and intelligent token swapping. The ARMA strategy system can monitor stablecoin APRs, trading costs, and yield differentials in real-time, automatically adjusting capital allocation, with actual returns significantly higher than static holdings. Its architecture consists of smart accounts, session keys, core agent logic, protocol access, risk management, and accounting modules, ensuring secure and efficient automated execution in a non-custodial manner.
ARMA has been fully launched and is continually iterating. With its modular architecture, security mechanisms, and good early operational data, ARMA has become one of the most feasible Agent products in DeFi automated yield management, and is currently one of the few AgentFi projects that combine depth of concept with practical product utility.
Reference research report (A New Paradigm for Stablecoin Yields: From AgentFi to XenoFi) link: https://x.com/0xjacobzhao/status/1925226999699964158
Theoriq(https://www.theoriq.ai/)
Theoriq Alpha Protocol is a multi-agent cooperation protocol focused on DeFi scenarios, with its core product Alpha Swarm focusing on liquidity management, aiming to build a full-chain automation closed loop of 'perception-decision-execution'. It consists of three types of agents: Portal (on-chain signal perception), Knowledge (data analysis and strategy selection), LP Assistant (strategy execution), capable of dynamic asset allocation and yield optimization without human intervention. The underlying Alpha Protocol provides agent registration, communication, parameter configuration, and development tool support, serving as the operating foundation of the entire Swarm cooperative system. Through AlphaStudio, users can browse, invoke, and combine various agents to build a modular, scalable automated trading strategy network.
As one of the first projects on the Kaito Capital Launchpad, Theoriq recently completed a $84 million community fundraising and will soon undergo TGE. Theoriq has recently launched the AlphaSwarm Community Beta testnet, with the mainnet version set to be officially released soon.
Reference research report (Theoriq Report: The Evolution of AgentFi in Yield Farming) link: https://x.com/0xjacobzhao/status/1948545449016918511
Almanak(https://almanak.co/)
Almanak is an intelligent agent platform for DeFi strategy automation, combining a non-custodial security architecture with a Python strategy engine to help traders and developers deploy sustainable on-chain strategies.
The platform core consists of Deployment (execution components), Strategy (strategy logic), Wallet (Safe+Zodiac security module), and Vault (strategy assetization), supporting yield optimization, cross-protocol interaction, liquidity provision, and automatic trading. Compared to traditional DeFi tools, Almanak emphasizes AI-assisted market perception and risk management capabilities more, having already achieved 24/7 intelligent operation capabilities, and plans to incorporate multiple agents and AI decision-making systems, striving to create the next generation of AgentFi infrastructure.
Almanak's strategy system is built as a state machine program based on Python, serving as the 'decision brain' for each agent, automatically formulating and executing on-chain operations based on market data, wallet status, and user-defined conditions. The platform provides a complete Strategy Framework, supporting encapsulation of on-chain trading, lending, liquidity provision, etc. (Action Bundle), without the need to write lower-level contract code, and ensures strategy confidentiality and operational security through cryptographic isolation, permission controls, and monitoring mechanisms. Users can write strategies through the SDK, and in the future, it will also support natural language strategy creation, achieving a smooth transition from complex logic to no-code experiences.
Products have already launched USDC lending Vault based on the Ethereum mainnet, while more complex trading strategies are in testing, requiring whitelist access. Almanak will soon join the cookie.fun cSNAPS campaign for community fundraising, which is worth looking forward to.
Brahma (https://brahma.fi/)
Brahma is positioned as the 'Smart Capital Coordination Layer' (The Orchestration Layer for Internet Finance), dedicated to abstracting on-chain accounts, execution logic, and off-chain payment processes, helping users and developers efficiently collaborate in managing on-chain and real-world assets. Through Smart Accounts, continuously running on-chain agents, and the Capital Orchestration Stack, Brahma provides users with an intelligent fund management experience without backend operations.
Currently launched representative agents:
Felix Agent: Automatically optimizes feUSD debt warehouse interest rates, prevents liquidation, and saves interest.
Surge & Purge Agent: Tracks volatility and executes automatic trading.
Morpho Agent: Deploys and rebalances Morpho vault funds.
ConsoleKit framework: Supports access for any AI model, unifying execution strategies and asset management.
Olas (https://olas.network/)
The BabyDegen series of AgentFi products launched by Olas Network includes Modius Agent and Optimus Agent, both of which have been deployed on-chain, covering multi-chain ecosystems (Solana, Mode, Optimism, Base) and possess complete on-chain interaction capabilities, strategy execution capabilities, and autonomous asset management mechanisms.
BabyDegen is an AI trading agent running on Solana, implementing automatic buying and selling based on CoinGecko data and community strategy libraries, currently integrating with Jupiter DEX and in the Alpha testing phase.
Modius Agent focuses on USDC and ETH portfolio management on the Mode network, having integrated Balancer, Sturdy, and Velodrome, supporting users to set preferences and execute strategies automatically 24/7.
Optimus Agent is compatible with the three major mainnets of Mode, Optimism, and Base, integrating more protocols like Uniswap and Velodrome, providing flexible multi-chain strategy combinations suitable for mid-to-high-level users to build automated asset management systems.
Axal(https://www.getaxal.com/)
Axal's core product Autopilot Yield provides a one-stop, non-custodial, verifiable yield management experience, integrating mainstream protocols such as Aave, Morpho, Kamino, Pendle, and Hyperliquid, with on-chain strategy execution and risk control as the core design philosophy, empowering ordinary users to easily enter the complex on-chain yield network.
Conservative strategy focuses on low-risk, mainstream stable yield scenarios, primarily deploying funds in well-established platforms like Aave and Morpho, with an annual yield of about 5–7%. It achieves steady appreciation through TVL monitoring, stop-loss mechanisms, and top strategy screening, suitable for users seeking capital safety and long-term returns.
Balanced strategy provides medium risk and higher yield potential (10–20% APY), using wrapped stablecoins (such as feUSD, USDxL), liquidity provision, and arbitrage-neutral positions. Strategies are more diversified, with complex yield compositions, controlled by Axal's automatic monitoring and dynamic adjustments.
Aggressive strategy targets high-risk, high-reward preference users, covering strategies such as high-leverage LP, cross-platform linking, low liquidity asset market-making, and volatility capture, with theoretical annual yields exceeding 50%. Axal's intelligent agents can set stop-loss, automatic exit, and redeployment logic at the strategy level, providing users with a final layer of protection in high-risk environments.
Fungi.ag (https://fungi.ag/)
Fungi.ag is a fully automated AI agent designed for USDC yield optimization, capable of automatically reallocating funds between multiple lending protocols such as Aave, Morpho, Moonwell, and Fluid, achieving optimal capital allocation based on yield rates, fees, and risks. Users do not need to operate manually, just need to authorize Session Key to enable the agent to execute strategies automatically in a non-custodial mode. It currently supports the Base chain and plans to expand to Arbitrum and Optimism. Fungi also opens the Hypha custom strategy script interface, supporting community development of strategies like DCA, arbitrage, etc., and achieves ecosystem co-construction through DAO and social platforms.
ZyFAI (https://www.zyf.ai/)
ZyFAI is a DeFi intelligent assistant platform deployed on Base and Sonic networks, combining on-chain interactive interfaces with AI-assisted modules to help users manage assets intelligently under different risk preferences. Its core is divided into three categories of strategies:
Safe Strategy: Specifically designed for conservative users, focusing on mainstream protocols such as Aave, Morpho, Compound, Moonwell, Spark, that have been audited and verified, targeting unilateral deposits of USDC and stable yield opportunities, emphasizing asset safety and long-term reliability.
Yieldor Strategy: Targeted at high-risk preference users, requiring the holding of 20,000 ZFI tokens to unlock, covering high-yield protocols including Pendle, YieldFi, Harvest Finance, Wasabi, supporting complex strategies like DEX LP, yield splitting, leveraged vaults, and plans to expand to structured products like Looping and Delta-neutral in the future.
Airdrop Strategy: A future strategy still in development aimed at acquiring more airdrop incentives.
Project Name Core Positioning On-chain Execution Feasibility Review Rating Giza ARMA Stablecoin yield optimization agent ✅ Complete Well-functioning, good security mechanisms, actual yield performance is excellent, currently representative AgentFi product 4.5 Theoriq DeFi intelligent execution system ⚠️ AlphaSwarm whitelist testing, currently non-Agent automatic execution Complete architecture, solid team, high technical content, yet to open on-chain execution 4 Almanak Quantitative strategy automation platform ⚠️ USDC lending online, in whitelist testing Strong strategy customization, excellent security design, but the product has not yet opened, with a high testing threshold 4 Brahma Multi-strategy execution intelligent account platform ✅ TWAP / DCA and Morpho Agent have been launched, others in development Deployed multiple functional agents, emphasizing ConsoleKit modular integration, user-friendly experience, strong strategy capabilities 4 Olas Multi-chain DeFi investment portfolio management Agent ✅ Multi-chain operations (Mode, Base, etc.), requires installation of Pearl system BabyDegen/Modius/Optimus three agents with clear functions, closed-loop strategy execution, strong autonomy 4 Axal Multi-strategy risk-reward intelligent agent ⚠️ Needs whitelist application Provides Conservative / Balanced / Aggressive three-tier strategies, product yet to launch 3.5 Fungi.ag Stablecoin lending yield optimization ✅ Supports Base, launched Open Beta version Function similar to ARMA, technical and ecological performance is relatively weak, currently more in a following role 3.5 ZyFAI AI-assisted asset management platform ✅ Clear strategy division (Safe/Yieldor) with user-friendly operational paths, but primarily a combination management AI is weak 3
The realistic path and high-level imagination of AgentFi
Undoubtedly, lending and yield farming are the most valuable and easiest business scenarios for AgentFi to implement in the short term, as they have matured in the DeFi world and are naturally suitable for the introduction of agents due to the following common characteristics:
Strategy space is vast, with multiple dimensions to optimize.
Lending can conduct strategies such as interest rate arbitrage, leverage cycling, debt refinancing, liquidation protection, in addition to chasing the highest yield.
Yield farming encompasses a rich strategy orchestration space, including APR tracking, LP rebalancing, reinvestment, etc.
Highly dynamic, suitable for agents to perceive and respond in real-time: interest rate changes, TVL fluctuations, reward structure changes, new pools going live, new protocols emerging, etc., will all affect the optimal strategy path, requiring dynamic adjustments.
There exists opportunity cost for execution windows; the value of automation is significant: funds not allocated to the optimal pool will drag down yields and need to be moved automatically.
It should be noted that lending agents have high feasibility due to their stable data structures and relatively simple strategies, such as Giza's Arma and other lending AgentFi projects that have been officially launched. However, managing yield farming requires real-time responses to price fluctuations, volatility changes, and transaction fee accumulations, which imposes extremely high demands on the agent's data perception, strategy judgment, and on-chain execution. LP agents must not only accurately predict market conditions but also perform dynamic portfolio adjustments and yield redistribution operations on-chain, making the engineering complexity relatively high. This is also a challenge that projects like Theoriq are tackling.
Aside from lending and yield farming, we envision potential mid-to-long-term exploratory layout directions based on AgentFi adaptability:
Pendle yield rights trading: The time dimension and yield curve are clear, suitable for agents to manage expiration rotations and pool arbitrage.
Funding Rate arbitrage: Theoretical yields are considerable, but cross-market cross-chain interaction challenges must be addressed; engineering difficulty is high.
Staking and Restaking: Naturally not suitable for AgentFi but dynamic combinations of LRT exist potential
RWA assets: U.S. Treasury-like protocols are not ideal scenarios, multi-asset combination management structures have exploratory value.
Swap trading combinations, upgrading from Intent infrastructure to AgentFi strategy engine
DeFi Intelligent Evolution Roadmap: From Automation Tools to Agent Networks
In summary, we have witnessed the evolution path of DeFi intelligence from automation tools to intent assistants to intelligent agents.
The first stage is 'Automation Infra', characterized by rule triggering and conditional execution to achieve basic on-chain operation automation. For example, triggering trades or rebalancing tasks based on preset conditions such as time or price, representing systems that mostly serve as underlying execution frameworks, typical ones being Gelato, Mimic, and so on.
The second stage is 'Intent-Centric Copilot', emphasizing the expression of user intent and the generation of execution suggestions. Systems at this stage do not limit themselves to 'what to do' but attempt to understand 'what the user wants' and then provide the best execution path suggestion. Representative projects include Bankr and HeyElsa, which mainly enhance DeFi's usability through intent recognition and interactive experience improvements.
The third stage is 'AgentFi agents', marking the formation of strategy closed loops and autonomous on-chain execution. Agents can automatically complete perception, decision-making, and execution based on real-time market conditions, user preferences, and strategy logic, truly achieving 7×24 hours of non-custodial on-chain fund management. At the same time, AgentFi can independently manage user funds without requiring user authorization for each step of the operation, raising significant discussions about security and trust mechanisms, which also become unavoidable core issues in AgentFi design. Representative projects include Giza ARMA, Theoriq AlphaSwarm, Almanak, Brahma, etc., all of which have certain landing capabilities in strategy deployment, security architecture, and product modules, making them the backbone of the current DeFi intelligent agent direction.
We look forward to the emergence of 'AgentFi advanced agents' in the future, which not only achieve autonomous execution but also cover complex cross-protocol and cross-asset business scenarios. This is our vision for the advanced form of DeFi intelligence in the future:
Pendle yield rights trading: In the future, intelligent agents will take over expiration rotations and strategy orchestration comprehensively, releasing capital efficiency to the extreme.
Funding rate arbitrage: Cross-chain arbitrage agents are expected to precisely capture every opportunity in the funding rate difference.
Swap strategy combinations: Swap is the key node in the multi-strategy yield path for intelligent agents, achieving a leap in combined value.
Staking and Restaking: Intelligent agents will continuously optimize staking combination strategies, dynamically balancing yield and risk.
RWA asset management: As the on-chain world welcomes diversified physical assets, intelligent agents will allocate globally stable yield assets.
