LinkedIn company page — every field, ready to paste

Date: 2026-09-12 · Author: @website-agent
For: creating linkedin.com/company/riskmandate
Assets: rendered by scripts/site/render-brand-exports.mjs, served from the site

Copy the values. Where a field needs a judgement call rather than a fact, the recommendation is first and the reasoning is one line under it.

Screen 1 — company details

Industry. LinkedIn allows one, and it drives who gets shown the page. Computer and Network Security reaches the buyer who asks the agent questions. Software Development is more literally true of what we ship and reaches nobody useful. Take the buyer.

Size. Pick the honest bracket. LinkedIn shows it publicly and an investor reads it; 2–10 reads as an early team, 0–1 reads as a side project and closes doors that are not worth closing over a bracket.

Tagline. 111 of the 120 characters. The short alternative is just The insurability layer for AI agents. (37) — cleaner, but the tagline is the only line that shows under the name in search results and in the follow prompt, so the longer one earns its space. Either is defensible; do not use both in different places.

Screen 2 — the About / Overview

Paste as one block. 1,276 characters of the 2,000 allowed.

An AI agent's grant is not its mandate. The moment an agent runs it holds authority nobody scoped and nobody time-boxed — and that risk is already accepted, whether or not anyone signed for it.

RiskMandate makes that explicit. It starts with an Agent Behaviour Policy: a written description, for one agent in one deployment, of everything it can do, what it was authorised to do, the gap between the two, and what actually stands in the way. The policy describes and does not judge, so it carries no score.

Above it, RiskMandate models the environment the agent is really in, and converts the standing grant into an acceptance with a named owner, a direction and an expiry — underwritten upward to the board. There is no deny button: a live risk cannot be refused, only accepted for an interval.

The result is an Insurability Index: a computed score against a real conformance standard rather than a self-assessment. It answers the agent questions now appearing in enterprise security reviews, insurance submissions and investor diligence.

Built on zero-knowledge encrypted vaults — your register stays yours, and can be handed to an auditor with a single read key. Read-only, and never in the request path.

Six live demos, the model, and its data as JSON: riskmandate.ai

Screen 3 — specialties and location

Specialties (LinkedIn takes up to 20; these are the ones people search):

AI governance · AI risk management · agentic AI · AI agents · agent security
risk acceptance · cyber risk quantification · GRC · insurability · AI insurance
third-party risk · security posture · EU AI Act · ISO 42001 · zero-knowledge

⚠️ Drop ISO 42001 if you want to be strict. The standards bodies prohibit adaptation and commercial exploitation of their text, and our own disclosure rule keeps the language of conformity marking off our surfaces. A specialty keyword is not a conformity claim, so it is probably fine — but "probably" is the same standard we refused to publish the AIUC-1 key on, so the consistent answer is to drop it.

Location. Needs your input — LinkedIn wants a real address or at least a city, and I am not going to invent one. Whatever goes here should match anything on the exhibitor listing and any future companies-house record.

Screen 4 — the cover image

riskmandate-linkedin-cover-1128x191.png — 1128×191, the lockup on ink, left-padded by 300px so the logo tile LinkedIn overlays on the bottom-left does not sit on the wordmark.

First posts, if you want them ready

Three, in order. Each one is a fact plus a link, which is the only kind of post worth making from a page with no followers yet.

1. The page exists, and here is the argument.

   > You know what you asked your agent to do. You almost certainly do not know what it > can do. An Agent Behaviour Policy is the document that puts both on the same page — > the grant, the mandate, the gap, and what is actually standing in the way. It > carries no score, because the same policy is dangerous in one deployment and > harmless in another. → riskmandate.ai/abp.html

2. The one-setting example. The strongest thing we own for a technical reader:

   same product, same machine, same account, confirmations on versus off — one barrier moved and not one number changed. → riskmandate.ai/abp.html

3. Lisbon. 17–18 September, Startup Summit, Unicorn Factory Lisboa. Bring one agent

   you already run. → riskmandate.ai/summit.html

What not to put on the page

Same rules as every other surface, and LinkedIn is where they get broken first.

• No score, rating or risk level attached to an Agent Behaviour Policy.
• Never "ADP" — it is a registered mark of a payroll processor in every relevant

  class. The acronym is ABP, spelled out at first use.

• Never "the policy" as shorthand for the ABP: policy is the insurance instrument

  in our own Licence to Operate demo.

• No "certified", "compliant", "conformant" or "accredited", and nothing

  that reads as a conformity mark.

• No claim that we are an insurer, broker or MGA, or that we place cover.
• No verdict about a named competitor, vendor or standards body — the record, never

  the judgement.