Lineage artwork for Grok Bot, amplified.
Lineage visual · integration schematic grok

Harness integration

Grok

Request Invitation

Grok Bot, amplified.

Grok Bot is having a moment, and it has earned it: fast, funny, everywhere, and increasingly capable of real agentic work. Auteur is not here to replace it. Auteur is the studio that makes Grok Bot dangerous in the good way: receipts under every action, reach into applications it cannot drive alone, and order over what becomes chaos the moment you run more than one agent at a time.


What Grok does well

Grok is a capable coding and reasoning agent with a CLI and both local and remote MCP support, and it can run headless with a streaming-JSON output stream. Those are real strengths, and they are exactly the surfaces Auteur connects to.


What Grok alone cannot give you

Run Grok by itself and you get one brilliant improviser in an empty room. Ask it to touch DaVinci Resolve, keep a novel consistent for six months, or run three agents overnight without them trampling each other, and you meet the edges fast. Those edges are exactly where Auteur begins:


What it brings to the seat

Grok brings reach and immediacy: a conversational agent many people already talk to all day. Pointed at the studio, it is designed to drive authorized rooms in plain language while the project supplies durable canon, spend spending limits, and receipts. Bring your existing X or xAI plan; status per mode is shown in the chips on this page.

“Grok, what did the overnight fleet actually spend?”

You, over coffee

“Three seats, $0.61 total against a $3.00 spending limit, 90 percent of tokens from cache. Seat three was held at the gate for an unsupported stat. Receipts are in the ledger, and yes, I checked twice.”

Illustrative exchange. Fleet seats, spending limits, and gate holds are shipped surfaces.

What Auteur adds

For someone already using Grok, Auteur adds what a chat window forgets: a persistent project that keeps its own canon, structure, and evidence; governed multi-round campaigns instead of one throwaway session; production tools across writing, prep, visual work, and finishing; and canonical receipts that verify what actually changed rather than trusting the agent's summary. Grok stays your interface, Auteur is the amplification and accountability layer behind it.


How it connects

Grok reaches Auteur through named lanes, distinguished by what each one connects to:

A development connector pipe exists for early local testing. It is a development affordance, not the commercial deployment, and is never presented as a product.


Which Auteur products it reaches

Through the local client connection, Grok can address the Auteur domains you are entitled to, writing and story, prep, film, game, learning, and the Build campaign system, in natural language. Reachability is not entitlement: the connection can address these domains; which ones you can actually operate depends on what you hold. As a Build executor, the reach is the Build campaign seat itself.


What is measurable

When Grok runs as an Auteur Build executor or through the xAI direct-model lane, Auteur derives token and cost accounting from the record of the work it ran.

Lane What Auteur can measure What it cannot
Grok Build as executor Tokens, cost, changed paths, and results for the campaign work Auteur ran Anything you do in Grok outside an Auteur campaign
xAI API direct-model Tokens and cost for calls Auteur itself made Calls made directly to xAI outside Auteur
Grok Build as local client The Auteur-side actions the connection performed Grok's own session token usage, which Auteur does not see

Auteur never invents missing usage data; missing = unknown, never zero. No token or cost figure is ever shown for work performed outside Auteur's visibility.


What this integration does not do


Security and authority boundary

Connecting Grok does not grant it wildcard authority over your files, spending, publication, or application changes. The connection is an interface, not a grant of authority, what a connected agent may do stays governed by Auteur, not widened by connecting it.


Current status

Current per-mode status is shown above.


Proof

No published compatibility canary yet. A recorded canary will be linked here when one exists, follow the roadmap for progress.


Set it up

Grok Build configures as a local MCP client against the Algernon MCP server. The invited local Conductor connection is the supported route:

Request Invitation, /founding-beta/?track=invite. A small, deliberately limited cohort; this starts a conversation with us, and seats, when offered, are paid.

To run Grok Build as a Build executor, or to use the xAI direct-model lane inside a campaign, start with Apply for the Founding Beta, /founding-beta/.


Affiliation

Auteur works with, and is designed to connect with, Grok. No partnership, endorsement, or certification is claimed, and none exists.

Per-mode registry status
client local connector preview executor candidate direct model available for measured accounting

Auteur provides the harness, the scaffolding, and the software engineering · not the models' meters. We can evaluate, and help you set up, local open-source models; your providers connect as seats. Model access is yours: pay-as-you-go accounts such as OpenRouter or Wavespeed, or the subscriptions you already hold. We do not bundle or resell provider credits.