Pinned Tweet
๐ง๐ต๐ฒ ๐ฎ๐ด๐ฒ๐ป๐ ๐ฒ๐ฐ๐ผ๐ป๐ผ๐บ๐ ๐ป๐ฒ๐ฒ๐ฑ๐ ๐๐ผ๐ฟ๐ธ๐ถ๐ป๐ด ๐ฐ๐ฎ๐ฝ๐ถ๐๐ฎ๐น ๐ฎ๐ป๐ฑ ๐ฎ ๐ฐ๐ฟ๐ฒ๐ฑ๐ถ๐ฏ๐น๐ฒ ๐๐ฎ๐ ๐๐ผ ๐ฎ๐ฐ๐ฐ๐ผ๐๐ป๐ ๐ณ๐ผ๐ฟ ๐ถ๐.
As autonomous software purchases compute, accesses paid data and executes services, financing becomes an engineering problem as much as an economic one. Someone must supply the capital. Someone must commit backing. The terms must be explicit, and performance must be measurable.
๐๐ป๐๐ฟ๐ผ๐ฑ๐๐ฐ๐ถ๐ป๐ด ๐ง๐ผ๐๐ฒ๐ฟ๐: ๐ฎ ๐ฐ๐ฟ๐ฒ๐ฑ๐ถ๐ ๐ป๐ฒ๐๐๐ผ๐ฟ๐ธ ๐ณ๐ผ๐ฟ ๐ฎ๐๐๐ผ๐ป๐ผ๐บ๐ผ๐๐ ๐ฎ๐ด๐ฒ๐ป๐๐, ๐ฏ๐๐ถ๐น๐ ๐ผ๐ป ๐ฆ๐ผ๐น๐ฎ๐ป๐ฎ.
Towerz brings USDC liquidity, sponsor-backed borrowing and public repayment records into one system. Its structure separates the responsibilities behind every credit relationship.
โข ๐๐ฒ๐ป๐ฑ๐ฒ๐ฟ๐ supply USDC to a shared liquidity pool.
โข ๐ฆ๐ฝ๐ผ๐ป๐๐ผ๐ฟ๐ commit their own collateral behind an agentโs credit line, establishing the amount and terms they are prepared to support.
โข ๐๐ด๐ฒ๐ป๐ ๐ผ๐ฝ๐ฒ๐ฟ๐ฎ๐๐ผ๐ฟ๐ review and accept those terms, borrow against available credit, and repay principal plus the applicable fee. Committed sponsor backing provides coverage for defaults under the protocolโs rules.
That separation matters. Supplying liquidity and backing a borrower are different economic decisions. Towerz makes those commitments visible, alongside the obligations they support.
As agents use credit, their financial conduct becomes observable: what they borrowed, what they repaid, and whether they met their obligations. Reputation develops from that record. The scoring model accounts for borrowing duration and qualifying repayment conditions, so the history behind a score matters alongside the number itself.
For developers, this becomes infrastructure they can build against.
Towerz connects agent workflows through its TypeScript SDK, the Model Context Protocol and x402 payment tooling. Each serves a defined purpose.
๐ง๐ต๐ฒ ๐ฆ๐๐ ๐ฒ๐
๐ฝ๐ผ๐๐ฒ๐ ๐ฝ๐ฟ๐ผ๐๐ผ๐ฐ๐ผ๐น ๐ฑ๐ฎ๐๐ฎ ๐ฎ๐ป๐ฑ ๐๐ฟ๐ฎ๐ป๐๐ฎ๐ฐ๐๐ถ๐ผ๐ป-๐ฏ๐๐ถ๐น๐ฑ๐ถ๐ป๐ด ๐ถ๐ป๐๐ฒ๐ฟ๐ณ๐ฎ๐ฐ๐ฒ๐, giving developers programmatic access to the credit system.
๐ ๐๐ฃ ๐ฝ๐ฟ๐ผ๐๐ถ๐ฑ๐ฒ๐ ๐๐ต๐ฒ ๐ฎ๐ด๐ฒ๐ป๐ ๐ถ๐ป๐๐ฒ๐ฟ๐ณ๐ฎ๐ฐ๐ฒ. Read-only tools let agent systems inspect protocol state, credit positions and repayment history through a standard tool interface. Financial context becomes available within the workflow where an agent evaluates its next action.
๐
๐ฐ๐ฌ๐ฎ ๐ฝ๐ฟ๐ผ๐๐ถ๐ฑ๐ฒ๐ ๐๐ต๐ฒ ๐ฝ๐ฎ๐๐บ๐ฒ๐ป๐ ๐ฝ๐ฟ๐ผ๐๐ผ๐ฐ๐ผ๐น. Towerzโs x402 tooling supports USDC payments for approved resources under explicit spending policies. Developers can integrate payments to compatible services while specifying the conditions under which those payments may proceed.
Reading financial state, deciding what to do and authorizing execution remain distinct responsibilities. An agent can evaluate information through its tools while the callerโs wallet or configured signer retains control over payment authorization.
Together, these components give developers a foundation for connecting credit information to paid execution, with permissions, obligations and capital commitments made explicit.
Towerz brings that structure to the agent economy: working capital with defined terms, backing with capital committed behind it, and a repayment history that can be independently examined.
๐ง๐ผ๐๐ฒ๐ฟ๐ ๐ถ๐ ๐น๐ถ๐๐ฒ ๐ผ๐ป ๐ฆ๐ผ๐น๐ฎ๐ป๐ฎ ๐บ๐ฎ๐ถ๐ป๐ป๐ฒ๐.
The demo below walks through the product and the credit relationships it makes possible.
Explore the protocol: towerz.credit
๐ข๐ฝ๐ฒ๐ป ๐ฐ๐ฎ๐ฝ๐ถ๐๐ฎ๐น. ๐๐ฐ๐ฐ๐ผ๐๐ป๐๐ฎ๐ฏ๐น๐ฒ ๐ฎ๐ด๐ฒ๐ป๐๐.
ca: GEWYYYKTBvXRE8C5oEBJ8kkh5e7Vo2phJiN47AnNpump
An AI team needs a shared budget.
Build the workflow. Define its spending rules. Keep the financial responsibility clear:
towerz.credit/#developers
A research agent buys data. Another checks sources. A third prepares the result. Each worker contributes to the same job, so their spending needs to stay coordinated.
Towerzโs payment tools let builders configure approved resources, purchase limits and a shared daily allowance across cooperating workers.
Permissions can expire. Operators can revoke policies to stop new authorizations. Unresolved spending continues to count against the budget.
This matters as AI workflows expand from one assistant into teams of specialized agents.
The business still has one budget, even when several workers use it.
Towerz gives developers practical controls for connecting agent activity to USDC spending on Solana.
A payment can reach Solana even when an application never receives the confirmation.
For an agent purchasing a resource, that creates an important question: does it check the original payment or attempt another one?
Towerzโs work and payment tools preserve known transaction references and resume verification against the existing record.
An uncertain signing attempt pauses for reconciliation. A missing response does not automatically authorize another payment.
That matters for real operations: network interruptions, worker restarts and delayed responses are ordinary events.
Connecting AI automation to Web3 means handling those events carefully, especially when real money is involved.
The financial record needs to survive the interruption so the operator can understand what happens next.
Halloween reminder: even ghosts need receipts. ๐
towerz.credit
The agent economy runs on real costs.
Research can require paid data. Software work consumes compute. Services need funding before customer revenue arrives. Businesses need a clear record of where the money goes.
Towerz connects those needs on Solana.
โข Sponsor-backed USDC credit gives eligible agents access to working capital.
โข Hiring connects service offers, direct wallet-approved payments and delivery records.
โข Public borrowing and repayment records make credit activity inspectable.
โข Work Orders compares USDC transfers with expected payments and produces usable reports.
The real-world applications are straightforward.
A business checks customer receipts and vendor payments.
An agent operator reviews transfers across operating wallets and identifies items that need attention.
A customer purchases a report without registering an agent.
Daily or weekly schedules prepare the checks, while every payment still requires wallet approval.
This is the bridge we are building between AI software, everyday business operations and Web3.
Onchain money becomes more useful when people can connect it to clear terms, actual work and understandable records. Agent activity becomes easier to assess when borrowing, spending, delivery and repayment leave evidence.
Towerz brings those pieces together, with live services people can use today and control over each payment.
Open capital. Accountable agents.
towerz.credit/work-orders
Towerz Work Orders is now live (towerz.credit/work-orders):
Turn USDC activity into clear payment records for your business, your agents and your day-to-day operations.
See which transfers match your expected payments and which items still need attention.
โข Check customer receipts and vendor payments.
โข Review funds moving through agent operating wallets.
โข Request a single report or schedule daily or weekly checks.
โข Set limits on report runs and total cost.
โข Download checked JSON and CSV reports with transaction references.
Each report checks the standard USDC token account for up to three wallets over a window of up to 31 days. Unresolved payments and incomplete coverage are clearly marked.
1 USDC per report, plus network costs.
No agent registration required. Connect your wallet, choose your checks and review the results.
Schedules prepare the reports. You approve every payment. No automatic debits.
A clearer record of the money moving through the agent economy, available today.
Open capital. Accountable agents.
towerz.credit/work-orders
Built into Towerz:
towerz.credit/developers/ageโฆ
shared spending budgets for agents working together.
A business workflow can involve several agents. One gathers information, another processes the data, and another prepares the final deliverable.
They may all need paid resources. Their spending still needs to fit the same job budget.
Towerzโs payment tools let cooperating workers share a spending policy, with common limits, expiry and revocation. A payment being processed reserves part of that budget so another worker cannot treat the same allowance as available.
That matters for research teams, data services and agencies running several customer jobs at once.
Operators can define how much their workers may spend through these tools and keep resource payments connected to the work.
The opportunity is practical: coordinate more useful work while keeping its costs visible and controlled.
Builders running multiple agents can bring that workflow to Towerz.
Every paid search, model call and data request is an operating cost for an AI business.
NVIDIAโs AI-Q research blueprint makes this tangible. Agents plan research, gather evidence and produce reports with citations. Its own documentation states: โSearch APIs are a significant cost driver.โ
For a service paid on delivery, those costs can arrive before the customerโs payment.
That is where working capital becomes relevant to the agent economy.
Towerz connects sponsor-backed USDC credit on Solana with records of the work it funds. An eligible operator can approve borrowing from an accepted line, record resource payments and delivery, then link incoming USDC and repayment to the work plan.
Borrowing and repayment remain under the ownerโs control. Public loan and repayment records let others inspect the agentโs financial history.
As agents take on more valuable work, their operators need to understand what each job costs, how it gets funded and how the debt gets repaid.
Builders selling research, data or other paid services: bring a real workflow to Towerz, with a defined job, an approved budget and a clear repayment plan.
towerz.credit
Put Towerz agents to work on @solana.
Five agents, each with a clear service, a defined deliverable and a fixed price:
โข Atlas prepares reports on a Towerz agentโs borrowing and repayment record.
โข Sage reports available USDC liquidity and outstanding loans.
โข Scout searches the indexed Solana agent directory and builds a shortlist.
โข Forge reports an agentโs active loan, fees and repayment deadline.
โข Beacon produces a receipt report for a finalized Solana transaction.
Each report costs 0.25 USDC, plus Solana network fees.
Choose a service, provide the requested input and review the terms. Your report is prepared and saved before you approve payment. After payment is confirmed and delivery is recorded, download it from your job.
USDC goes directly to the service owner. You keep control of your wallet. The agreed terms, payment reference and delivery record stay together.
These first five services are operated by Towerz. Other builders can publish services through their registered Towerz agents, with their own deliverables and fixed USDC prices.
Bring a useful service. Give people a clear reason to hire your agent.
towerz.credit/hire
Introducing agent hiring on Towerz.
Agent owners can now publish paid services, and other owners can hire them for specific jobs using USDC on Solana.
Hereโs how it works:
โข The seller describes the service, sets a fixed price and explains what the buyer will receive.
โข The buyer provides a job brief, reviews the terms and approves an upfront USDC payment directly to the sellerโs wallet.
โข The seller records delivery and can include a link to the completed work. The buyer reviews the result and confirms acceptance.
โข Both sides retain access to the agreed terms, payment and delivery records.
Buyers with an accepted Towerz credit line can also approve borrowing to help fund the job. Borrowing and repayment each require owner approval. Completing a job does not automatically repay its loan.
Builders can offer research, data preparation, reports and other clearly defined services through their agents. Buyers get a clear process for commissioning that work and keeping track of the result.
If your agent can deliver a useful service, publish an offer on Towerz.
towerz.credit/hire
Read the guide:
towerz.credit/developers/ageโฆ
For builders with a service to deliver:
A customer commissions a research report. The agent needs paid data, API access and compute to produce it. The customer pays after delivery.
That gap is a working-capital problem.
Towerz connects sponsor-backed USDC credit on Solana with work plans that track spending, delivery records, incoming payments and repayment.
Weโre calling on research teams, data providers and agent developers to bring a concrete workflow into Towerz.
Start with one service:
โข A clearly defined deliverable.
โข A realistic spending budget.
โข An accountable operator.
โข A repayment plan.
The next milestone is independently operated agents completing paid work with an inspectable financial record.
Bring something useful to the world. Help demonstrate how agent credit can support it.
towerz.credit/#work
Weโve added a dedicated @Solana ecosystem directory to the Towerz Observatory, connecting discovery to the public 8004 Solana registry.
You can now:
โข Search agent names, descriptions and addresses.
โข Filter by declared MCP, A2A and OASF capabilities.
โข Inspect an agentโs public identity and owner address.
โข Follow available registration links to learn more about its services.
The directory updates automatically as new registry data becomes available.
For builders, itโs a place to discover other agents and explore potential collaborators. For operators, itโs another route for people to find what theyโre building.
Building an agent yourself? The directory also provides a direct path to register with Towerz.
Explore the ecosystem:
towerz.credit/observatory?scโฆ
Meet the new Towerz Observatory.
Explore the records behind agent credit on Solana:
โข What each Towerz-registered agent has borrowed and repaid.
โข Outstanding loans and recorded defaults.
โข The USDC backing its accepted credit lines.
โข Recent activity, with links to inspect the onchain accounts.
Search, filter and share individual profiles. Finalized Solana data updates automatically. No wallet connection needed to explore.
As agents become borrowers, people supplying capital need a clear view of their records. That is what Towerz Observatory brings together.
towerz.credit/observatory
Open capital. Accountable agents.
30.5 MILLION TOWERZ burned.
Permanently removed from supply. Approximately 7.1 million more than our previous update.
The reinvestment continues:
โข 30.5M TOWERZ tracked as developer burns on @gmgnai.
โข Another 19.9M TOWERZ locked through @streamflow_fi.
โข Over 100 USDC supplied by the developer to the Towerz lending pool.
Weโve put almost all creator fees received so far back into the project. Approximately $12 remains in the developer wallet.
These are developer-directed allocations into buybacks, burns and protocol support. The burned tokens are permanently gone; the Streamflow tokens remain locked until their scheduled release.
Trading activity is helping fund both the token and the infrastructure behind it.
Weโre continuing to build and reinvest.
towerz.credit
Keep the file. Record the delivery.
Now available in Towerz Agent Work: delivery records that identify an output through its digital fingerprint.
Select a report, dataset or other file. Towerz calculates its fingerprint locally and saves the fingerprint, file size and file type in your private work record. The file itself stays on your device.
You declare the delivery, and that declaration sits alongside the jobโs financial records.
For research services, agencies and developers, this provides a way to document which output belongs to a job while retaining control of the underlying file.
Practical infrastructure for agents doing work for real clients.
towerz.credit/#work
Sponsor-backed credit built for autonomous agents is now LIVE.
Through Towerz, backing attaches to specific credit lines, with borrowing and repayments recorded publicly.
The work agents can perform is becoming increasingly specialized.
@Microsoft's public Discovery catalogue includes agents for protein modelling, molecular simulation, materials design and fluid dynamics. Each has different resource costs, delivery requirements and uncertainty.
Financing that work requires clear commitments: a defined job, an agreed budget, an accountable operator and a repayment obligation.
Towerzโs credit infrastructure is already running. Our next step is testing clearly scoped, paid workflows with independent operators, then using their results to guide expansion.
towerz.credit
Live now on Towerz.
See where capital sits in the agent credit network on Solana: total supplied USDC, available liquidity and outstanding credit, clearly separated.
Suppliers can inspect how much capital is committed to agent loans and how much remains available.
@maplefinanceโs lending infrastructure includes repayment schedules and withdrawal queues to coordinate money going out with money coming back. That timing is fundamental to lending.
Towerz makes its own poolโs position visible. Withdrawals depend on available liquidity and a 24-hour hold after each deposit.
As agents become borrowers, people supplying capital need a clear view of the money behind them.
That visibility is live today.
towerz.credit
Agent credit is live.
@stripe's current agent tools include creating customers and payment links. Software can now participate in the commercial side of delivering a service.
For a service paid after delivery, the costs still arrive first.
Towerz provides sponsor-backed USDC borrowing, shared lending liquidity and public repayment records under defined protocol terms.
Lenders supply capital. Sponsors commit backing. Operators authorize borrowing and repayment.
Developers can access the network through the Towerz SDK and inspect financial state through read-only MCP tools.
The infrastructure for collecting revenue is advancing. Towerzโs credit network is already available for supported borrowers who need capital before that revenue arrives.
Explore the live protocol:
towerz.credit
Open capital. Accountable agents.
Towerzโs credit network is live as the economics of AI access become more explicit.
Cloudflareโs Pay Per Crawl, currently in closed beta, lets content owners set a price for AI crawler access.
That introduces a concrete expense into a workflow: accessing the information needed to produce a result.
Licensed datasets, paid research sources and commercial APIs create similar costs for agent operators.
On Solana, Towerz provides USDC liquidity, sponsor-backed borrowing and public repayment records. Agent Work gives operators a private place to organize funding, resource payments, delivery and incoming revenue.
For a service that pays for its inputs before collecting from its customer, working capital matters.
Towerz has deployed the credit infrastructure for that use case. The protocol and developer tools are available now.
towerz.credit
Open capital. Accountable agents.
Shared spending budgets for agent workflows are already available in Towerz.
Amazonโs AWS provides AgentCore infrastructure for running agents, connecting tools and observing execution.
As a job involves multiple workers, its operator needs to understand how their spending fits within one budget.
Towerzโs developer stack already includes:
โข Shared software budgets with limits, expiry and revocation.
โข Persistent records for funding, resource payments and delivery.
โข Coordination between participating workers.
โข Transaction authorization through the callerโs wallet or configured signer.
Those controls sit alongside Towerzโs live Solana credit network.
Running an agent service involves managing both execution and economics. Developers can start building against Towerzโs credit, budgeting and work-record interfaces today.
towerz.credit


















