Governed agents

Governed agents with a budget and a run history

An agent with no governor is a rumor about what happened to your data. Here every Oxygent, OXYGEN's in-app AI agent, and every custom agent caps each paid step and records what it did.

try
Start freeRead the docs

On the ladder from most to least predictable, Tables with AI columns come first and take the most setup, then Workflows. Agents, Oxygents included, are the easiest to turn on and the least predictable, so the platform enforces their limits instead of leaving them to a prompt.

An Agent run's published version carries its authorization. Full access to the eligible capability catalog exists only after an explicit standing acknowledgement, narrowed by optional allow, ask, and deny rules. In an Agent run, tool authority is never spend authority. A manual run has an approved credit ceiling, and a cron, event, or webhook trigger stores the same ceiling per delivery.

Example searches

Ask for it the way you'd say it

Each example is a request in plain words, with what it returns underneath.

  • A worker for the inbox

    draft a reply for every new inbound conversation but never send one without me

    The drafting agent is held to draft-only by a database constraint. A draft goes out when you send it, or when an Oxygent or a Full-access Agent sends it.

  • Ask before anything external

    let it research freely but stop and ask me before any message or paid call

    Agent runs pause per ask rule at a checkpoint you resume; an Oxygent chat on Ask for approval holds sends and larger paid steps.

  • Read what it actually did

    show me what ran last night, what it cost, and where it stopped

    A run lists status, tool calls, approvals, inference usage, credits, and a deep-link to whatever it produced.

Plays

Three motions you can run this week

Each one is a chain of OXYGEN primitives — the same hosted objects your workspace already has, composed.

  • Turn on a job instead of building one

    A recurring chore runs continuously without anyone authoring the mechanism behind it.

    1. Agents
    2. Messages
    3. Observability
    1. 1Pick a shipped agent whose owning primitive matches the job you want done.
    2. 2Enable it, read the first run end to end, and only then leave it alone.
    oxygen agent enable

  • Give it a schedule and a ceiling

    Unattended work happens on a cadence with a per-delivery spend cap you approved in advance.

    1. Agents
    2. Workflows
    3. Observability
    1. 1Create a cron or event trigger on the Workspace Agent and set the credit ceiling it carries per delivery.
    2. 2Review the run list weekly rather than the transcript daily.
    oxygen agent trigger-create

  • Keep it grounded in your own knowledge

    Output reads like your company because the run pulled your positioning, not a generic prior.

    1. Knowledge Graph
    2. Agents
    3. Records
    1. 1Fill the typed core so every run injects the same profile, offers, and voice.
    2. 2Let learnings file back as sources, so next month's runs start better informed.
    oxygen agent run

Capabilities

What you get

  • Limits that hold in every mode

    Platform deny rules, each step's credit cap, and your balance live in the runtime, not in agent instructions. Auto-approve skips approval cards, never these.

  • Immutable versions

    An Agent run's authorization, context policy, and skill snapshots are pinned to a published version, so the run is explainable by something fixed. An Oxygent chat reads its Oxygent's current settings at every message.

  • Every action on the record

    An Agent run keeps its status, credits, tool calls, and checkpoints on the run; an Oxygent chat keeps each action and approval in its event log.

  • A roster with no placeholders

    Everything listed runs today: your Oxygents and specialists for reply drafting, meeting notes, knowledge synthesis, and inbound tagging.

Boundaries

What Agents does not own

Agents own goal-directed runs, not the objects they produce. A drafted reply is a Messages row, a filed note is Records activity, a synthesized page belongs to the Knowledge Graph, and deterministic child orchestration is Workflows. Uploaded runtimes are never executed here.

Data sources

Data sources and CRM connections

  • Exa

    Grounding search for a run that needs a fact the workspace does not already hold.

  • Pa

    Parallel

    Deeper research for the jobs where the assignment is a question rather than a lookup.

  • Serper

    Plain search results, often all a qualification step needs before it moves on.

  • Firecrawl

    Reads the page an agent chose to check, shown as a tool call on the run.

  • People Data Labs

    Person and company records for a job that has to resolve who somebody actually is first.

Use managed credits where available, or connect a supported provider key on an eligible paid plan. Your provider bills its calls; OXYGEN adds a per-call platform fee. Signed-in CRM connections such as HubSpot have no per-call fee. Connected sending accounts reserve monthly credits. See pricing and supported integrations.

Limits

Where this stops

  • An agent is the least predictable rung on the ladder. If you can write the steps, a workflow does the job more cheaply and predictably.
  • Platform deny rules and hard ceilings are not configurable. That is reassuring when something goes wrong and constraining when you want an agent to do something the capability catalog does not expose.
  • Cron, event, and webhook triggers belong to the Workspace Agent, your first Oxygent; specialists keep their own cadence. An Oxygent with a published version can run recurring work as a routine set up from its chat, with per-run and daily credit caps.

FAQ

Questions people ask first

Which agents actually ship today?
Your workspace's first Oxygent opens on Home, and you can add more. Beside them run 4 specialists: inbox reply drafting, a meeting notetaker that files notes onto records, knowledge synthesis on a cadence, and an inbound conversation tagger.
Can an agent spend without asking?
Yes, within limits. An Agent run spends inside its approved ceiling, per manual run or per trigger delivery. Your Oxygent chat starts on Auto-approve, so paid steps run without a card, each within its own credit cap and your balance.
Can I build my own?
Yes. Add an Oxygent and give it its job in chat, or create a custom agent with its own instructions, tool access, and versions. OXYGEN never runs a runtime you upload.
What stops it sending something embarrassing?
The published posture and your chat's approval mode. Reply drafting is draft-only, an ask rule pauses an Agent run before its call, and full access needs sign-off. Your Oxygent chat starts on Auto-approve, which also answers ask rules; switch to Ask for approval to hold sends, sequence starts and larger paid steps.
How do I debug a bad run?
Open the run. It carries status, the tool calls it made, approvals it waited on, inference usage, credits drawn, and links to the artifacts it wrote, so you diagnose behavior rather than reconstruct it from output.
Is this an agent builder?
Not for agent code. You can add Oxygents and custom agents with their own instructions and tool access, but they run over OXYGEN's bounded capability catalog. If you want a framework for agent code, this is the wrong product and we would rather say so.

Run governed agents with a budget and a run history on your own workspace

Sign up, get a working workspace with a one-time credit grant, and run the first play from the web app, the CLI, or your AI assistant over MCP.

Start freeSee pricing