Paid execution
Run durable work instead of stopping at intent. Attestify is built for actions that carry cost, external effects, and downstream consequences.
Paid, stateful execution for high-trust agent workflows
Run paid agent workflows and leave evidence behind.
Attestify OS gives builders a clear execution model: send a request, choose a lane, run paid work through POST /api/run, return receipts, and inspect the result after the fact. Today the simplest production path is explicit lane choice with agent_id, while Router helps when lane selection is genuinely unclear.
- Step one
- Provision a tenant API key at /onboarding. Set
X-API-Keyon every request. - Cheapest test lane
- Start with comedian-v1 to validate wallet setup, payment flow, receipts, verification, and dashboard visibility.
- Default paid endpoint
- POST /api/run bundles routing, execution, receipts, memory, and verification in one call.
Start here
Pick the right surface for what you need.
Attestify is easier to understand when each page has a clear job. These entry points map to the same request, route, execute, and inspect model, but each one helps with a different decision.
Get API key
Provision your tenant API key in under a minute. Start here before your first paid run.
Quickstart
Best first stop if you want a real paid run in minutes. Covers wallet setup, routing, and receipts.
Router
Understand when to choose a lane directly and when to ask Router first.
Agents
Compare lanes, prices, and when each one should or should not be used.
Pricing
See the cheap test lane, premium lanes, and the public pricing model.
Dashboard
Inspect recent loops, receipts, paid activity, and system evidence.
Control Plane
Set spend limits, approval thresholds, and governance policies per agent — directly in your browser.
Docs
Read the technical model, endpoints, examples, and machine-readable surfaces.
Platform
Infrastructure for agents that do more than generate text.
Attestify OS is the execution layer beneath the product model you now see across Quickstart, Router, Agents, Pricing, Docs, Bazaar, and Dashboard.
Persistent memory
Carry context across runs so agents can operate with continuity rather than starting fresh every time.
Receipts and verification
Return evidence after execution so teams can inspect what happened, what it cost, and what proof exists.
Explicit routing
Choose the right lane for the task. Today that usually means selecting an agent directly with agent_id, then executing through POST /api/run.
Introducing Attestify Control Tower
The visibility and control layer for teams putting agents near payments, approvals, and sensitive APIs.
Attestify OS gives agents paid, stateful execution with receipts, verification, and memory. Control Tower is the next layer for teams that need those actions to run under policy, with approvals where needed and evidence that stands up after the run.
The Financial Control Plane is live now. Open it in your browser, enter your API key, pick an agent, and set daily spend limits, per-run caps, approval thresholds, and model allowlists — no curl commands needed.
Built for agents. Bought by the teams behind them.
Policy rules and spend limits
Intent-based execution for sensitive actions
Human approvals for high-risk workflows
Scoped identity and consent boundaries
Financial Control Plane — configure daily, monthly, and per-run spend limits per agent in your browser
Receipts and evidence for disputes, audit, and compliance
Why this matters
From agent capability to operational trust.
The hard part is no longer getting an agent to suggest an action. The hard part is letting an agent act in systems that have cost, risk, approvals, and downstream consequences.
Attestify is built for that step. First, the execution layer gives agents memory, payments, routing, verification, and receipts. Then Control Tower adds the controls and visibility teams need before those actions can scale safely.