BUILD auto+gate brief-from-gap

Every content gap becomes a writeable brief

The file

take it and run it.

One markdown file: the goal prompt, the verifier prompt, and the stopping condition. Read it before you run it. MIT licensed, no account, nothing to sign up for.

Download brief-from-gap.md View the raw file

curl -fsSL https://vibemarketers.ai/loops/brief-from-gap.md -o .claude/loops/brief-from-gap.md

The loop

copy it.

Finding a content gap is the easy part. The gap sits in a report until someone turns it into something a writer can actually start from.

Use it when you have a topic map or a keyword gap list and nothing has been written against it.

You need the gap list and a claims file, so briefs cannot cite things you cannot prove.

You get one brief per gap: the question, who asks it, what the page must cover, and which proven claims it may use. Gaps belonging elsewhere are flagged rather than written.

The gate is you approving briefs. The loop never writes the page.

Goal prompt

Read the topic map for <company> and take every cluster marked as a gap. For
each, write a brief naming the question it must answer, who is asking it,
what the page has to cover, and which proven claims it may cite. Mark any
gap that belongs to a different property rather than this one. Never write
the page.

Verifier prompt

Open every brief. FAIL if a brief exists for a gap that is not in the topic
map, if a brief cites a claim that is not in the claims file, if a brief has
no audience, or if the loop wrote a page rather than a brief. Otherwise PASS.

Stopping condition

Every gap in the topic map has exactly one brief, or a reason it was skipped, and the verifier marks PASS.

MCPs it needs

none required

Want everything at once

The installer writes the three starter skills and every published loop into .claude/. Read the script first.

curl -fsSL https://vibemarketers.ai/install.sh | bash

Or have us run it on your site →

The receipt

Receipt brief-from-gap 2026-09-25

Baseline: A topic map with 11 unanswered buyer questions and no mechanism to turn any of them into something a writer could pick up.
Hypothesis: A gap plus the claims file is enough to write a brief automatically, because the question sets the scope and the claims file sets what may be asserted.
Intervention: Read the topic map, took every remaining gap, and wrote a brief naming the question, the audience, what must be answered and which claims may be cited.
Result: 1 gap read and 1 brief written, after a questions page closed the other 10. The brief was marked as belonging to SuperMarketers rather than the lab, because it asks how work follows through to revenue.
Commercial signal: none
Attribution limits: A brief is not a page and this loop has not shown that briefs written this way produce good pages. It ran against a single remaining gap, which is too small a sample to claim the format works.
Learning: The loop is only as useful as the gap list feeding it. Running it after the gaps were closed meant it had almost nothing to do, which is the correct behaviour but a thin first run.
Next intervention: Run it against a topic map with real gaps, on a site that has not just closed them, and check whether the briefs are specific enough to write from.

0 human touches

3 minutes

0 errors

raw run logopen →

Sources: run log, lab/runs/2026-09-25-brief-from-gap.json, collected 2026-09-25 , window: 1 gap, single run .

One entry per run, append-only, newest first. Corrections append a dated line.

Inside a company

where the lane changes.

VibeMarketers tests the loop. SuperMarketers runs it on a real team.

Problem: Content gaps get found and then sit in a spreadsheet because nobody turns them into something writeable.
Source of truth: The topic map and the claims and proof file.
Permitted actions: Read gaps, write briefs, flag out of scope gaps. Never write a page, never invent a claim.
Required approvals: The content owner approves each brief before anyone writes against it.
Owner: Content lead
Metric: Share of gaps with a brief, and share of briefs that became pages.

On a real team the brief carries the sales context behind the question, so the writer knows which objection it is answering.

Running this on your own site is one person reading a diff. Running it inside a company changes four things, and the guide covers them. When it stops being a file you copy and becomes a build, SuperMarketers does the same loop model inside your stack, on your keys.