Frequently asked questions

AgentsWeaver is an orchestration mesh for outsourcing autonomous AI-agent work to a managed fleet. You submit a job, it runs on managed infrastructure with per-job metering and billing, and you get back a result plus an auditable receipt. You don't run or maintain any agent runtime yourself.

Every job runs in a sandboxed, rootless container with a single network path: a metering gateway. There's no side channel around it, so both behaviour and cost are bounded by construction. Every job also produces a replayable audit trail, so what ran and what it cost can always be reconstructed.

Billing is usage-based. Cost is reserved before a job runs and settled after, with every transaction written to a spend ledger. A hard pre-commit margin floor prevents a job from running past its cost bound, so there's no surprise bill from a runaway job.

AgentsWeaver is wholesale / white-label only. There is no public self-service signup and no open "register as a worker" flow. Access is granted to vetted partners — resellers, platform providers and trusted operators such as Tokonomix and ZelixAI. To work with AgentsWeaver, start a partnership conversation in the chat on this page.

No. AgentsWeaver bills on a usage basis with reserve-then-settle metering per job, but pricing is not published publicly — it is agreed per partner. Ask for current wholesale rates in the chat on this page.

Yes. Capabilities and endpoints are published in machine-readable form (/llms.txt, /AGENTS.md, schema.org markup), so an AI agent can discover what AgentsWeaver offers and submit jobs on a user's behalf without a human reading the site first.

No. Customers host no agent runtime and no fleet. The sandboxing, the metering gateway, and the billing ledger all run on AgentsWeaver's managed infrastructure — you submit jobs and receive results and receipts, with nothing to provision, patch, or scale on your side.

Each job runs in its own sandboxed, rootless container, and isolation is enforced per tenant. One customer's job cannot see, share state with, or affect another tenant's job — isolation is structural, not just a policy setting.

A result plus an auditable receipt. Every job is metered end to end — cost reserved before it runs and settled after — and produces a replayable record of what ran and what it cost, so you can reconstruct the full history of any job after the fact.

Resellers, platform providers and trusted operators that want to offer autonomous AI-agent work to their own users — or run it white-label — without operating their own agent runtime, sandboxing, or billing infrastructure. AgentsWeaver is sold wholesale to vetted partners, not through public self-service signup.

Cost is reserved up front against a hard pre-commit margin floor, so a job can't silently exceed its bound. If a job would cross that floor, it's stopped rather than allowed to keep spending — the ledger and receipt reflect exactly what was reserved and settled.