CoreCMO

Your Operator Brief


Fill this once. Every prompt in the playbook will speak to your real company.

This is the central context file. Once you've populated it, every prompt card on the site will substitute your real company name, your real ICP, your real voice rules: no more [COMPANY NAME] placeholders. Your Brief has three states, and you control the move between them: it autosaves to this browser as you type; once you're signed in it also syncs to your CoreCMO account so it follows you across devices and to anyone you invite; and it stays private to you and your team until you explicitly promote it to a published company page, which is the point at which any of it becomes public. Nothing publishes on its own. Be honest in here: a Brief you softened is a Brief that teaches AI the softened version.

Why this document matters more in 2026

"AI helps amplification. It does not inherently generate impact. If what you offer and why is fuzzy, AI multiplies the fuzziness. So before you multiply with AI, ask: what are you multiplying?"

This Brief IS what you're multiplying. Every AI tool you point at marketing work, every prompt, every agent in AI Operating Model (Web Operations, Performance Marketing, Field Marketing and the rest), every output across the playbook: runs on this context. Sloppy Brief, scaled wrongness. Tight Brief, scaled rightness. Fill it well.

Status

No Brief saved yet. Fill the fields below: autosaves as you type.

First time here? Paste your company URL below and the four research agents fill the Brief in 10 seconds. Edit anything before it saves.

Every field is labeled: Public data AI estimate: refine Yours Still needs you

AI populate

Generate your Brief

Paste your URL. The research agents fill your Brief in about 10 seconds.

Have your own assistant fill the rest

The site read above can only know what your website says. Your own assistant can see your CRM, your wiki and your deal reviews, so it can answer the questions a website never states. Paste the instruction below into it with this page open.

Worth knowing before you do: this page holds about a third of the Brief. The rest live inside the module that asks each question, which is also where the question is explained. The instruction points your assistant at a manifest of all of them, so it knows where to go.

Fill my CoreCMO Operator Brief. First, fetch https://corecmo.ai/assets/brief-fields.json. That is every field in the Brief, each with its label and the page it lives on. Fields marked /brief are on the page I have open. The rest are inside the module named, one page each. Then work through the fields you can answer from what you can actually see: our CRM, our wiki, our docs, past deal reviews, win and loss notes, our own product pages. For each one, type the value into the matching input and move focus away so the page saves it. Rules, in order of importance. 1. NEVER change a field that already has a value in it. Whatever is there is either mine or something I have already reviewed, and my answer beats your inference every time. Skip it and tell me at the end which ones you skipped and what you would have put instead. 2. NEVER fill a field marked "waiting on someone". That marker means a human has already looked at this field, knows the answer, and knows it is not theirs to give. It is the one kind of blank that is a decision rather than a gap, and filling it overwrites that decision with a guess. Report them to me as a list instead, because that list is who I need to go and ask. 3. Leave a field EMPTY rather than guessing. An empty field is a known gap and every agent reading this Brief treats it as one. A plausible guess is a wrong answer that forty downstream agents will quietly trust. If you are not sure, it goes on the list you hand back to me, not into the field. 4. Say where each value came from. When you are done, give me a table: field, the value you entered, and the source you read it from. Anything you inferred rather than read goes in a separate list marked as inference. 5. Do not publish, do not change account settings, and do not touch anything outside the Brief fields. Start with the page I have open, then tell me which module pages are worth visiting next and why, and wait for me before you go there.

If you have connected CoreCMO over MCP, prefer corecmo_record_finding instead. It reaches every field in one call rather than a page at a time, it refuses to overwrite an answer you gave, and it records which agent supplied each value so the Brief shows it as agent-sourced. Typing into the page does none of those three: a typed value is indistinguishable from one you wrote yourself, which is why rule 1 above has to be a rule rather than a guarantee.

Section 1 · TAM, SAM & Market Sizing


Does the market exist?

The market math every downstream area of the playbook depends on. Source: TAM, SAM & Market Sizing.

Section 2 · ICP & Audience


Who you sell to.

The customer-context layer every targeting prompt and agent inherits from. Source: ICP & Audience.

Section 3 · Buyer Personas & Propensity


The humans inside the ICP.

One persona per buying-committee role, with propensity-to-buy signals per role. Source: Buyer Personas & Propensity.

Section 4 · Right-to-Win


Where you win consistently: the empirical truth.

The bridge between knowing your ICP and being able to charge a premium for it. Source: Right-to-Win.

Section 5 · Pricing & Packaging


What you charge, and how.

The pricing philosophy, tier architecture, value metric, discount policy, competitive position. Source: Pricing & Packaging.

Section 6 · Brand & Positioning


Who you are.

The brand-context layer every prompt inherits from. Source: Brand & Positioning.

Visual brand assets : optional. Fill these and the Presentation Builder can produce fully-branded, board-ready decks with zero prompting.

Business complexion : these shape budget, ICP, and persona work to your actual situation instead of a generic default.

Section 7 · KPIs & Measurement


What you measure.

The metrics your CFO has signed off on. Source: KPIs & Measurement.

Section 9 · Norms & Decision Rights


What you are allowed to decide.

Facts tell an agent what is true. Methods tell it how to work. Neither tells it what it may decide without asking you, so it either guesses or stops. This section is the difference between an artifact you can send and one you have to rewrite.

Section 11 · Commitments & Capacity


What you have already promised.

Every prioritization question is really a displacement question: what comes out to make room. An agent that does not know what you committed to will happily tell you to do the new thing, because it has nothing to weigh it against. These four fields are what turn "should we" into "instead of what".

Saved


Your Brief is saved.

Every prompt card on the site will now substitute your real values for the bracketed placeholders. Try it.

Brand & Positioning ICP & Audience

Your fields autosave to this browser, and to your CoreCMO account once you're signed in, so the Brief follows you across devices. Clearing your browser data wipes the local copy; the account copy restores it next time you sign in. Clear saved Brief removes both and cannot be undone. It does not take down a company page you have already published: unpublish that from the publishing panel above.

Need to present this to your CEO or board?

Publish your Brief (below) and paste your private context link into your own Claude or GPT project. It carries the CoreCMO Operator Agents: ask it to "turn my Brief into a board-ready deck in our brand" and it builds a branded presentation from your real positioning, ICP, and metrics. The same context can generate any master deliverable and facilitate your strategy and team workshops.

Optional · LLM context distribution


Publish your Brief as corecmo.ai/c/your-company.

One URL. Three surfaces. Any AI agent, any LLM, any analyst can read your operating context.

  • /c/your-company: Schema.org-marked human page. Citable in LLM answers.
  • /c/your-company/brief.json: Machine-readable Brief substrate. AI agent ingest.
  • /c/your-company/llms.txt: Per-company AI agent manifest.

Private fields (KPIs, pipeline data, named customers) are stripped automatically. You control the slug, you control publish/unpublish, you control republish. Auto-republishes on Brief save once a slug is claimed.

Sign in to claim your slug.   Sign in

Section 8 · AI Voice Rules


How AI represents you.

The specific rules every agent in your stack must follow. Reinforces Sections 6.11–6.13.

Section 10 · Company Facts


The facts an AI will repeat about you.

Auto-populated from public sources on every run. Correct anything that is wrong: these travel into your context file and into every answer an assistant gives about your company.