Features

GrailAOS features, each with its status label

Each feature carries the status label that says how much of it runs today and, when needed, the limit that changes a decision.

Status labels

How to read the status labels

A statement is a short, verifiable sentence with an id that never changes. Its status label says how much of it runs today.

How to read the labels

Runs today
Implemented and tested; if it ships turned off, the limit says how to turn it on. If the statement is not a capability, such as the license, the trademarks or a design principle, it means the statement is true today.
Partial
Runs, with gaps in the implementation: the limit says which ones.
Available on request
Does not run today: Omniatix™ implements it if a customer asks for it, with no timeline.
Coming soon
Omniatix has announced it and it is not available yet. No date.
Vision
The purpose GrailAOS™ is designed for. It is not a current capability: what already runs carries its own label.

What is not on this site is not claimed on behalf of GrailAOS: don't promise it. For other capabilities, ask Omniatix.

The limit
The limit that changes a decision goes next to the label: what it does not do yet, or how to turn it on if it ships turned off.
The verification
Each statement was verified against the code and the specification; those references are internal.
To cite
Quote the sentence with its label and link to the page where it appears.

Verified against the code on Specification version implemented by the runtime v4.437 (measured on )

By area

Features by area

Each one, with its status label.

The language

  • Runs todayA solution in text documents
  • Runs todayLibraries for pieces that repeat
  • PartialSpecification versions

    Limit A minor version can change behavior, and there is no automatic migration between major versions.

See more about The language

The compiler

  • Runs todaySix stages before anything runs
  • Runs todayDiagnostics written for a model
  • PartialFixes that get checked again

    Limit The assisted loop runs only on the command line.

See more about The compiler

Testing

  • Runs todayTests inside the solution
  • Runs todayEverything external simulated, an honest report
  • Runs todayNo false green

    Limit At the gate, failing cases warn by default; blocking is declared in the solution.

See more about Testing

Availability and scale

  • Runs todayMultiple , one turn at a time

    Limit The per-conversation queue and the shared turn queue are turned on in the 's configuration.

  • Partial as the source of truth

    Limit If is lost, the deferred jobs held there are lost; the no-Redis mode is turned on by configuration.

  • Runs todayStandard , today on
See more about Availability and scale

Message delivery

  • Runs todayWhatsApp delivery that does not lose messages
  • Runs today in integrations

    Limit Off by default, and turned on tool by tool.

See more about Message delivery

Releases and live changes

  • Runs todayReleases without rebuilding the image
  • Runs todayPreview and
  • Runs today by moving a pointer
See more about Releases and live changes

Workflows and scheduled tasks

  • Runs todayWorkflows that run on their own
  • Runs todayA person's approval
  • Runs todayRuns that continue on another server
See more about Workflows and scheduled tasks

AI models

  • Runs todayClaude, with failover

    Limit The failover is off by default and is turned on together with the operator alerts.

  • Runs todayBounded determinism
See more about AI models

Extensibility

  • Runs today as a client
  • Runs todayDelegation between agents
  • Runs todayPlatform services

    Limit The deployment's startup code wires in contacts, opportunities and the calendar, and Google Workspace needs its own credentials.

See more about Extensibility

Channels

  • Runs todayWhatsApp, web chat, and voice devices
  • Runs todayPhotos the assistant sees

    Limit Turned on with the solution's vision setting; assisted photos only in web chat.

  • PartialVoice over the phone

    Limit On a single pod, and without handing off to a person during the call.

See more about Channels

People in the loop

  • Runs todayHandoff to a person
  • Runs todayAsking the team

    Limit Turned on by configuration; the person replies from the console.

See more about People in the loop

Memory

  • Runs todayMemory of what each customer shares

    Limit Off by default; turned on in the deployment's configuration.

  • Runs todayReminders a person asks for
  • Runs todayEach conversation's history, in Postgres
See more about Memory

Security and tenants

  • Runs todayIsolation between
  • Runs todayIntegrations without private network access
  • PartialOperator permissions

    Limit It covers only some actions, and some skip the check if the console does not send the permissions.

See more about Security and tenants

Observability

  • Runs todayThe of every turn
  • Partial and

    Limit OpenTelemetry sends one span per agent turn: workflow runs are not sent.

  • Runs todayAlerts to the operator
See more about Observability

Usage and caps

  • Runs todayReal cost per turn
  • Runs todayUsage caps and alerts

    Limit The turn in progress always finishes; usage notifications are off by default and, like the billed-amount cap, need billing turned on.

See more about Usage and caps

The operations console

  • Runs todayThe console screens
  • Runs todayTeam, roles and sign-in
  • Runs todayRuns and approvals in view
See more about The operations console

Enterprise deployment

  • Runs todayDedicated deployment
  • Runs todaySign-in with the company's identity
  • Runs todayData on the customer's infrastructure
See more about Enterprise deployment

Context-aware support

  • Runs todaySupport that sees the screen and can watch it live with permission

    Limit Turned on by configuration.

  • PartialSupport in the dedicated console

    Limit With sign-in through the company's identity provider, it does not hand off to a person yet.

See more about Context-aware support

LLM-first design

  • Runs todayA question behind every decision
  • Runs todayThe loop until it compiles and passes its tests
  • Runs todayBuilt with coding agents
See more about LLM-first design

Access, license and trademarks

  • Runs todayProprietary engine
  • Runs todayPrivate developer program
  • Runs todayPatents pending
See more about Access, license and trademarks

On request

What is available on request

It does not run today: Omniatix implements it if a customer asks for it, with no timeline.

  • Available on requestOther providers, through an OpenAI-compatible
  • Available on requestCalling another solution
  • Available on requestTelegram as a channel
  • Available on requestManaged Kubernetes: Amazon EKS, Google GKE or Azure AKS
  • Available on requestA dedicated connector for a company system

More information

What you've read is less than half.

What you see on this site is a summary. A company evaluating GrailAOS can get the rest, with the detail and the limits of each capability. It isn't published: we share it with whoever asks for it.

shown here
65
already running
more than 180

Behind them is a specification of 1,385 specified features and 564 architecture decisions, measured on October 30, 2026.

Is your company evaluating GrailAOS? Request the rest

Leave your details, with your company email, and we'll write to you with more information about GrailAOS.

Up to 500 characters.

The request goes to the Omniatix team by email.

For machines

The same statements, for machines

The same content, with its status labels, in formats a program can read.

Runtime The engine that runs a program that is already written. In GrailAOS, the one that operates the compiled solution.Pod The unit in which Kubernetes runs one copy of a service: adding pods means adding copies that serve at the same time.Deployment A specific installation of a system, with its own configuration and data.Postgres A widely used relational database that stores data durably.Redis A very fast in-memory store, used as a cache and to coordinate several copies of a service.Kubernetes The standard system for running containers across several servers: it spreads the work, restarts what fails and updates one copy at a time.k3s A lightweight distribution of Kubernetes, with the same standard interface.Idempotency Repeating an operation has the same effect as doing it once: a retry does not duplicate a payment or a message.Canary A new version that first reaches only part of the users, so it can be seen working before everyone gets it.Rollback Going back to the previous version when the new one is not working well. As a verb, roll back.MCP Model Context Protocol: an open protocol that lets a model use the tools and data that another server offers.API An interface that lets another system use a service through requests and responses, without going through a screen.Tenant and multi-tenant A tenant is an account, for example a business, with its data kept apart; multi-tenant means that several of them share the same installation without seeing each other's data.Trace The record of what happened in a turn: what was done, which tools were called, how long it took and how much it cost.Prometheus A standard system that collects a service's numeric metrics to chart them and raise alerts.OpenTelemetry An open standard for sending traces and metrics to whatever observability tools each company uses.Gateway An intermediary that takes requests through a single interface and passes them on to different providers.