Idea

Agents that react to the world

Mailboxes, inbound event sources and standing work orders, so an agent team runs on what happens and never only on prompts

Event-driven agentsThree systemsMailboxes · event sources · standing work
Events, at any hour A standing work order, open-ended

An agent that only answers prompts is a tool. An agent team that runs a business has to react to things that happen while nobody is typing: a payment lands, a customer replies, a post gets a comment, a document arrives, a job finishes somewhere else. I built the agent platform so that every seat has a mailbox, every external system can become a governed event source, and a work order can stay open indefinitely as a standing lead that turns events into finite tasks and brings specialists online when they are needed.

The shape

Every team lead and every seat has a mailbox. Seats can message only their own team, and the identity on every message is derived on the server from the authenticated participant, so a message cannot claim to come from someone else. The terminal result of an asynchronous tool call is delivered to the caller's mailbox, so messaging is part of governance rather than a side channel beside it.

External systems register as event sources against a work order. The lead turns each event into a finite task under the same rules as any other work.

One inbound event, end to end
1a provider signs a webhookoutside2signature checkedthe edge3persisted, then batchedthe edge4delivered across hostshosts5in the lead's mailboxthe work order6a finite taskthe work order
When delivery is uncertain
The batch is retainednever dropped, never blindly replayed

Receiving an event causes nothing at the provider.

A work order can have an open-ended envelope, so continuous operation is composed from the same machinery that governs a bounded engineering program, with no new primitives.

Bounded program

Tasks planned up front, seats staffed for the plan, and an end.

Standing envelope
LEAD

A standing lead admits a task when an event arrives and brings a specialist online only while it is needed.

Where it appears

In AgentOS, mailboxes, inbound event sources, standing-envelope work orders and event-driven tasks are part of the control plane, with the same work graph, hooks and review behind them: Why long agent runs fail.

In FAFOHive, the governed gateway's signed, durable return path carries a payments, email or social-publishing provider's events to the mailbox of the team lead whose assignment caused them, and three binding tools let a team bind the websites and mailboxes it owns and nothing else: Agents with real tools and no keys.

In Agent Swarm, agents message one another over durable streams with signed envelopes, so an agent that is busy or restarting loses nothing addressed to it, and in-flight work from an agent that stops heartbeating is reclaimed.

What it buys and what it costs

It buys agents that can run a process rather than answer a question: a business, an operations desk, a long-running integration. It buys attribution for every reaction, because the event, the task and the result are all nodes in the same graph. It buys resilience, because a delivery that is uncertain waits rather than disappears.

It costs an edge that must check and persist before it delivers, and a standing lead that must be governed as carefully as a bounded one. Both are in the platform.

The judgment in it

The decision was to make events first-class work. The alternative, a side channel that pokes an agent, would have given me reactive agents without governed ones.

An event that becomes a task inherits budgets, evidence and review for free.