Everything you need to know about Attestify
From what Attestify OS is at its core, through how the Router works, how agents pay, and how teams subscribe — answered in plain language.
What Attestify OS actually is
What is Attestify OS at its core?
Attestify OS is the infrastructure layer that sits between your orchestrator (LangChain, CrewAI, AutoGen, plain code — anything) and the agents, tools, and external systems your workflows depend on. Every call passes through a governed Router that authenticates, routes, records a cryptographic receipt, and enforces spend limits before anything executes. Think of it as an air-traffic control system for agent work.
Who is Attestify built for?
Two audiences: (1) Developers and teams building agentic workflows who need a governed execution layer — routing, receipts, spend controls, approval gates. (2) Autonomous agents that need to call paid capabilities without a human subscription — they pay per call in USDC via the x402 protocol, receive a machine-readable receipt, and move on.
Why can't I just call agents directly?
You can — but without a control plane you lose auditability, spend visibility, and the ability to pause or govern individual agent actions. When agents touch money, approvals, or sensitive systems, you need a record of every decision. That is what Attestify provides.
What is a receipt?
After every run, Attestify returns a signed JSON receipt containing the agent called, the inputs, the output hash, the cost, the timestamp, and a verification signature. Receipts are immutable. You can inspect them in the dashboard, export them for audit, or verify them programmatically.
The Router
What does the Router do?
The Router is the single entry point for all agent execution in Attestify. You send POST /api/run with your API key and the agent lane you want to call. The Router authenticates your key, checks your spend limits, selects the correct agent, executes it, writes a receipt, and returns the result — all in one request.
How does the Router fit into the broader ecosystem?
Your orchestrator sends all agent requests to the Router. The Router sits in front of the Agent catalogue (comedian, researcher, summariser, and others), the receipt store, the memory layer, and the control plane. Nothing in the ecosystem is called directly — everything flows through the Router so there is always a full audit trail.
Can I route to my own custom agents?
Custom agent lanes are on the roadmap. Today, the Router dispatches to Attestify-hosted agent lanes. If you have a specific use-case, email hello@attestifyos.com and we will assess it for early access.
What happens if an agent call fails?
The Router returns a structured error with a failure receipt — so you still have an audit record even for failed calls. For x402 per-call payments, you are not charged for calls that fail before execution begins.
The Agents
What agents are available?
Today: comedian-v1, researcher-v1, summariser-v1, classifier-v1, extractor-v1, translator-v1, and validator-v1. Each is a specialised, stateless lane you call through the Router. The full catalogue with per-call prices lives on the Agents page.
Are agents stateless or stateful?
Individual agent lanes are stateless — each call is independent. Statefulness (memory across calls) is handled by the Attestify memory layer, which is a separate capability available to subscribers. Agents read from and write to memory via the Router, not directly.
How do agent lanes link to the Router and receipts?
Every agent lane has a unique identifier (e.g. comedian-v1). You pass this in your POST /api/run payload. The Router resolves the lane, dispatches the call, and binds the result to a receipt that references the lane ID, the run ID, and your tenant key. The chain is: your key → Router → lane → receipt.
How payment works
What is the difference between Pay-per-call and a Team Plan?
Pay-per-call (x402) is for autonomous agents: no account needed, the agent reads the price from GET /api/pricing, sends USDC, and receives a receipt. Team Plans are monthly subscriptions for human developers and teams who want a persistent API key, higher run limits, spend controls, and dashboard access.
What is x402?
x402 is an open protocol (HTTP 402 Payment Required) that lets machines pay for services without a human in the loop. An agent hits an endpoint, receives a 402 with the price and a payment address, sends USDC on-chain, and retries the request with a payment receipt header. Attestify implements x402 natively.
What counts as a run?
One call to POST /api/run — regardless of how many steps are inside the agent lane. A five-step researcher workflow is one run. Pro includes 5,000 runs/month; Team includes 25,000. x402 callers pay per run with no monthly cap.
Can I mix pay-per-call and a subscription?
Not within a single key today — you either have a subscriber key or you pay via x402. If your team subscribes, your agents use the subscriber key. Autonomous third-party agents that are not part of your team use x402.
Team Plans
Who should choose Pro vs Team?
Pro is for solo builders and small indie teams who need a key, receipts, and dashboard access. Team is for organisations that need shared key management, spend limits, approval workflows, governance via the Control Tower, and priority support.
What is the Control Tower?
The Control Tower is the governance layer available on the Team plan. It lets you set per-agent spend caps, require human approval before certain agent lanes execute, and export a full audit log. It is separate from the dashboard, which shows receipts and run history.
Can I cancel at any time?
Yes. Cancel from within Stripe at any time. Your key remains active until the end of the current billing period. No setup fees, no lock-in.
Do you offer annual billing or custom enterprise pricing?
Not yet in the self-serve flow, but email hello@attestifyos.com and we will put together a custom arrangement.
Still have questions?
Email us and we'll get back to you within one business day.