Manifesto 2.0 pivots Warden from protocol-first to distribution-first. The App and Agent Hub becomes the primary surface for publishing, discovering, and calling AI agents.
WARD will launch in Q4 2025 with a 1 billion supply and serves as the token for network payments, validator bonds, incentives, and governance.
Chain support extends execution to Solana, Ethereum, Base, and BNB Chain. Verification and settlement remain on the Warden Chain while connectors let agents operate on destination chains.
New integrations and Agents extend Warden’s reach and capability. DeBridge packages cross-chain bridging behind confirmations, and Hyperliquid is slated to provide an AI-assisted trading terminal aligned with receipt-backed execution.
Testnet and Warden App Beta traction provides a baseline for distribution speed and usage. ~1 million WAUs in August and a Uniswap agent with 650,000 swaps from ~500,000 users in three weeks indicate early demand.
Introduction
In August 2025, Warden introduced Manifesto 2.0 and pivoted to a distribution-first strategy built around verifiable agent execution. Warden Protocol is an application-specific blockchain designed to embed AI into onchain applications and build a Global Agent Network. Its core verifier, Statistical Proof of Execution (SPEx), checks AI outputs probabilistically through sampling and validator consensus, yielding lower cost and latency with bounded false-accept and false-reject rates that operators tune through sample sizes and acceptance thresholds.
Following the manifesto, Warden began concentrating on user demand in the Agent Hub, a product embedded within the Warden App. Teams publish agents and resources, users discover them, and each call produces a receipt that downstream contracts can trust. Asynchronous Verifiable Resources (AVRs) supply reusable tools and data sources, Agent Passport anchors identity and portability, and Warden Studio covers building, testing, and monitoring. Early integrations extend capability and settlement venues so agent outputs can be consumed across multiple networks without bespoke pipelines.
The upcoming WARD token genesis will connect this distribution stack to governance and economics via fee payments, network security, incentives, and governance. Its design aligns incentives with receipt-backed demand, while supply and float choices target a market structure that rewards reliability and uptime over surface-level activity. As Manifesto 2.0 rolls out, Warden will scale verified Agent Hub usage, broaden resources and integrations, and hold verification costs to target error bounds before mainnet and the TGE.
Warden exists to make AI agents safe to use inside onchain applications. It standardizes how agents are published, discovered, and called, then turns execution into receipts that other contracts and apps can trust. The target users are teams that want agentic functionality without building their own verification, identity, and settlement stack. The tradeoff is lower cost and latency with bounded false-accept and false-reject rates that operators control through sample sizes and acceptance thresholds.
Blockchain Layer: The Warden Chain provides the base ledger and control plane for agents. Agents deploy to Warden Chain, sign requests, and pay and collect fees using onchain accounts. Each agent has a unique cryptographic identifier.
Verification Layer: The verifiability layer converts model execution into evidence that can be checked. SPEx samples and verifies that the invoked model produced the observed output and flags when results fall below reliability standards.
Application Layer: A set of tools to use, build, test, and publish agents to users. Warden, the application, is an everything app that allows agents to be used by millions of monthly active users. Warden Studio is the primary interface where teams can focus on agent logic while the protocol handles identity, verification, deployment, and monitoring.
Big Brain: A protocol-integrated, domain-specific language model under development. It trains on network interactions to improve agent performance and includes a mechanism to reward contributors who make it better.
At the core of Warden’s design are Asynchronous Verifiable Resources (AVRs), customizable AI plugins that bundle specific functions or data sources. For instance, an AVR might fetch token prices from an exchange API or run a model that predicts volatility, outputs that agents can reuse across chains. With IBC and Hyperlane support, AVRs can reach 100+ networks, positioning Warden as omnichain AI middleware. Agent identity and portability are formalized via Agent Passport, enabling the same agent to plug into multiple apps.
The Agent Hub is the primary product surface. Developers publish agents and resources into the Hub, users and apps discover them, and each call produces a receipt with inputs, outputs, and context. Receipts are the bridge between offchain inference and onchain settlement, so downstream contracts can act on results without bespoke integrations.
The end-to-end flow is straightforward. A user or app submits a task to an agent in the Hub, the agent executes using attached resources, SPEx verifies a statistical sample of calls and records receipts onchain, and if acceptance thresholds are met, results settle and fees route to the applicable parties. A verified call is a user or app request that passes SPEx checks and produces an onchain receipt. If accepted, it finalizes as a settled action, tying the input to a finalized, auditable outcome. Other apps can then consume those receipts as composable evidence.
SPEx verifies that the invoked model likely produced the observed output within a bounded operator-tuned error rate. It does not certify the global correctness of the model or its training data. When results fail thresholds, the system flags the call, with options to resample, escalate verification, or revert dependent actions based on policy.
Developers manage this lifecycle in Warden Studio. Teams attach AVRs for tools and data, write tests, and publish versions to the Hub. Studio exposes tracing, receipt previews, and reliability metrics so developers can tune prompts, resources, and sampling parameters before and after release. Coverage grows with added resources, increasing what agents can accomplish, and expanding chain connectors that allow receipts to settle cross-chain. Support for cross-chain connectors helps propagate receipts and payments across networks, while curated integrations in the Agent Hub concentrate discovery and reuse.
Warden is live in a limited configuration with core verification, early resources, and the Agent Hub. Near-term work focuses on expanding resource libraries, tightening SPEx parameters for common workloads, improving Warden Studio’s testing and observability, and adding first-class SDKs for partner apps.
Manifesto 2.0: Distribution First
Manifesto 2.0 reorders execution by leading with distribution via the Agent Hub, while routing every call through verification so each verified call and settled action produces an auditable receipt.
What Changed
Pre-Manifesto 2.0 plans emphasized core protocol components. The new plan prioritizes discoverability, onboarding, and monetization of the Agent Hub, then scales verification and transport for agents with repeat, receipt-backed usage. The benefit is faster iteration on real usage, measured by publish-to-first verified call and groups of repeat verified calls. The distribution loop extends into a Global Agent Network. Agent Passport anchors reputation, receipts, and portability so repeat usage can be ranked and monetized across venues.
The Agent Hub
The Agent Hub, embedded within their consumer-facing Warden App, is Warden’s marketplace and runtime for third-party agents. It serves as a directory for calling agents and a usage/ranking layer. By concentrating demand at a single entry point, Warden reduces the number of custom integrations and standardizes confirmation flows.
Manifesto 2.0’s distribution-first growth flywheel only works if verification guarantees hold, since users and developers will only compound adoption if results are consistently auditable. Warden’s loop secures this process through three steps:
Task: A user or app submits a request to an AI agent in the Agent Hub, which may invoke Asynchronous Verifiable Resources (AVRs) to package functions or data sources for offchain work.
Verified Call: SPEx verifies a statistical sample of executions and records an onchain receipt, establishing an auditable link between input and output.
Settled Action: Cross-chain transport moves requests and receipts to the chains where liquidity and users are active. This loop links interactions to verifiable outcomes and allows conversion to be measured at each stage.
In beta, the Uniswap Trading API Agent executed 650,000 swaps from ~500,000 users across 14 EVM chains in three weeks, averaging just over one swap per user and moving from deployment to production in under 72 hours. These results show how quickly the Hub can distribute agents, establishing a baseline for distribution speed and early usage that future listings can be measured against as more of the roughly 8.5 million testnet users come online.
Beyond trading and bridging, Warden has added agents such as:
Hyperliquid (Perp trading): An integrated Hyperliquid trading terminal is slated to provide a dedicated AI-assisted trading interface within the Agent Hub, aligning derivatives execution with the best signals and research information from the Messari Agent.
DeBridge (Cross-chain bridging): A dedicated bridging agent packages approvals, messaging, and settlement behind confirmations and explorer-visible receipts. Treating bridging as an agent pattern creates a template for any state-changing action requiring strong checks.
Messari (Deep Research): Generates Messari-style briefs and token sections from onchain and project data, standardizes terminology, cites sources, and emits a receipt per request for auditable output and ranking.
Jupiter (DEX routing): Routes swaps across liquidity sources to seek the best swap execution on Solana and records a receipt for each trade.
Levva (DeFi optimization): An automated portfolio manager that analyzes users' assets and provides and automates, and executes a DeFi strategy based on the users' risk tolerance
Kaibot (SocialFi): C Analyzes users social activity and mindshare leaderboard positions, generating a strategy to grow mindshare for different pre TGE projects on Kaito leaderboards.
Base Farmer (Airdrop farming on Base): An agent that allows users to accelerate transactions, swaps, and activity on Base to maximize airdrop eligibility
Travel Agent: An agent that discovers and books hotel rooms and flights based on the user's preferences and demands. Once confirmed, the booking is paid in cryptocurrency by the agent, and the user receives flight tickets and hotel confirmations via email.
Each integration increases either distribution or capability. Together, they reduce the need for tailored workflows per chain and let developers ship agents once and reach multiple networks through a single publish surface
Product & Adoption Updates
Recent releases focus on extending where agents execute, lowering the cost of packaging capabilities into AVRs, and proving that the call-to-confirmation loop holds under load. A task is the requested computation, a verified call is the receipt confirming execution via SPEx, and a Settled Action is the state change onchain.
EVM Support for Warden and the Agent Hub
Support for Solana went live on June 2, 2025, with Ethereum and Base following on July 11, 2025, and support for BNB Chain going live on Aug. 28, 2025. Support for these chains enabled chain-specific connectors and transaction surfaces so agents can read, write, and trigger transactions on the destination chain. Verification, receipts, and fee settlement still occur on the Warden Chain testnet. While per-chain confirmation times and Verified Call success rates are undisclosed, the team reported ~1 million weekly average users (WAUs) in August.
This expansion set the stage for broader distribution via the Agent Hub. The Agent Hub opened to third-party publishers on Sept. 10, 2025. Near-term health can be read through submission throughput, review times, and transparent rejection criteria. Tying ranking to verified calls and settled actions will determine whether discovery compounds utility rather than surface-level metrics.
Warden Studio and AVR supply
Warden Studio, launched with AVR marketplace support on July 11, 2025, is the developer console for building, testing, and publishing agents into the Agent Hub. AVRs package functions or data sources that agents can call asynchronously, and their reuse across agents is a key signal of modularity. Warden Studio enables streamlined submissions of agent listings and AVR packages, but counts of AVRs published, approval times, and reuse rates have not been disclosed. As these metrics become available, they will show whether Warden Studio is shortening the path from prototype to Agent Hub listing and driving network effects through reusable modules.
WARD Token Generation Event
WARD will launch in Q4 2025 with a fixed supply of 1 billion tokens. The initial supply distribution will be as follows:
Liquidity: 5% (50 million WARD) will be available at genesis to seed initial liquidity when WARD trading goes live.
Validators & Operators: 10% (100 million WARD) delegated to validators, but locked in perpetuity.
Public Goods: 10% (100 million WARD) for testnet incentives, genesis airdrops, and public-good initiatives. 19.4 million WARD will be available at genesis. The remaining allocation will experience accelerated vesting to 64 million within the first six months, and full vesting to 100 million over 30 months.
Ecosystem & Community: 12% (120 million WARD) allocated to community initiatives. 20 million WARD will be available at genesis, followed by a linear vest over 30 months.
Agent & Developer Incentives: 19% (190 million WARD) for agent, developer, and builder incentives. 20 million WARD will be available at genesis, followed by a linear vest over 12 months.
Core Contributors: 20% (200 million WARD) reserved for the core development team, advisors, and early contributors. This allocation is subject to a six-month cliff followed by two years of linear vesting.
To date, Warden has raised 450k in a single investment round in April 2024
Treasury & R&D: 24% (240 million WARD) for long-term treasury and research programs. 120 million WARD will be available at genesis, followed by a linear vest over 12 months.
This structure is designed to support builders and early adopters, while reserving significant resources for treasury management, ecosystem expansion, and governance participation. Following mainnet under Proof of Authority with zero emissions, the protocol plans to enable inflationary issuance upon transition to Proof-of-Stake (PoS), using a programmatic schedule that targets 65% staked WARD and adjusts between 1% and 10% annually, initially set at 8% and moving up when staking is below target and down when above.
Network Payment: WARD will be used as payment for fees on Warden Chain and Warden, for execution and verification costs, and for settling gas abstraction, so users will be able to operate across EVM-compatible environments without holding external native assets.
Agent Fees: Access to premium agents is priced in WARD.
App Subscriptions: A premium version of Warden is unlocked by staking WARD or via a monthly WARD subscription.
Publishing Agents: Developers pay a flat WARD fee to publish agents and use WARD to update agents and AVR packages.
Verification: All model verification services are payable in WARD.
Validator Bonds: To operate a validator, operators will be required to hold an undisclosed amount of WARD.
Incentives: A portion of WARD is planned to be used for initiatives that contribute to Warden Protocol's ecosystem and community projects.
Governance: WARD is planned to be used for community governance initiatives.
Together, these token utilities connect token demand to governance, network security, and usage across the Agent Hub. Notably, an undisclosed portion of Warden’s revenue is planned to buy back WARD. Once the token is live, tracking staked supply, fee revenue settled in WARD, premium feature adoption, and incentives per verified call will gauge protocol health.
Closing Summary
Warden is executing a distribution-first plan by centering the Warden App, its Agent Hub, and routing activity through the Task → Verified Call → Settled Action pipeline, which now runs across multiple environments with auditable receipts. App adoption suggests real interest, with hundreds of thousands of weekly active users and hundreds of thousands of swaps executed through agents such as the Uniswap Trading API, indicating that agents and AVRs may attract usage before verification and transport scale up. Partnerships and integrations, including DeBridge for bridging and Hyperliquid for trading, expand distribution and capability, positioning Warden as omnichain AI middleware rather than a chain bounded by its own ecosystem.
The upcoming WARD token launch ties network payments, validator bonds, incentives, and governance to verified usage so token demand reflects network security and activity across the Agent Hub. Readiness for mainnet will be signaled by disclosure of confirmation times, success and dispute rates, and curation quality. The near-term question is whether App traction converts into durable mainnet adoption and a healthy WARD economy anchored in receipt-backed usage, which can be seen by whether cohorts show repeat usage, fees begin to accrue, and staking participation rises.
This report was commissioned by Warden Protocol. All content was produced independently by the author(s) and does not necessarily reflect the opinions of Messari, Inc. or the organization that requested the report. The commissioning organization may have input on the content of the report, but Messari maintains editorial control over the final report to retain data accuracy and objectivity. Author(s) may hold cryptocurrencies named in this report. This report is meant for informational purposes only. It is not meant to serve as investment advice. You should conduct your own research and consult an independent financial, tax, or legal advisor before making any investment decisions. Past performance of any asset is not indicative of future results. Please see our Terms of Service for more information.
No part of this report may be (a) copied, photocopied, duplicated in any form by any means or (b) redistributed without the prior written consent of Messari®.
