01 /
What a posture document looks like
The posture document for a fictional Team Install on QM, redacted where the install fills in names, so the access table and the actions that wait for a person can be read before you order.
Notes
Written down, like everything else.
Short pieces on how agent deployments actually work, from hardening to the support promise. Each one derives from the same documents the whole site is built on, so nothing here contradicts a page you have already read.
02 /
Can this scale past the first team
A Team Install is set up as an organisation with scoped workspaces from the first day, so the second team is a line on the menu rather than a second install. What that line covers, and the one case that is a migration instead.
03 /
What hardened means
Least privilege by default and a posture document you sign. The word that does the most work on this site, unpacked.
04 /
Why we never host your agents
Optaimi has no hosting product, deliberately. What client-owned infrastructure means in practice and why it is safer.
05 /
Our SLA, read honestly
Response guaranteed, resolution best-effort. What each line of the support promise means and why the edges are published.
06 /
Choosing between Solo and Team
One person's workflow points to Solo, team-wide needs point to Team. How the recommendation is made when you answer unsure.
07 /
What an add-on is
A one-off module at $400, $700 or $2,000, built at the framework's extension points. The menu read closely, the two items that carry no price, and where it ends.
08 /
Maintain, Operate or Partner
Three retainer tiers, each containing the one below. How to read the ladder and which work sits inside each price.
09 /
What you sign at handover
Every deployment ends with a signed handover document. What it records, and where responsibility sits once the agents are yours.
10 /
What the first year costs
Every standard price is published, so the first year can be totalled before you speak to anyone. The arithmetic, and the one cost we will not estimate for you.
11 /
An agent is not a chatbot
A chatbot answers in its own window. An agent works in your tools, and everything else about this service follows from that difference.
12 /
Which tasks to automate first
The first task is rarely the most valuable one you can imagine. How the published menu, the default posture and the product choice narrow it down for you.
13 /
What prompt injection is
The risk that makes agents different from chatbots, what the default posture does about it, and why it sits on our published exclusion list rather than in small print.
14 /
Why not build it yourself
An answer to the objection that does not argue you could not. What the price is set against once the build is done, and when doing it yourself is the right call.
15 /
Is your system on the list
How to read the connector menu for your own stack: why the named systems are examples rather than a boundary, the connection you have already paid for, and what happens when your system is not named.
16 /
What ends when the engagement ends
An install closes at handover, so a retainer is the only part still running when you decide to stop. What ends with it, and why there is no credential of ours to withdraw.
17 /
Who owns what we build
A deployment can outlive its supplier and still be a trap. Where customisation lands, what happens when it reaches the framework core, and what you hold at the end of it.
18 /
When an agent acts on its own
The question sounds like it is about the agent, but the answer is a property of the action. What decides which actions wait for a person, who decided it, and what it does not settle.