Flags what your sign-off is likely to catch - before it does.
You write a newsletter, a post, a page. It goes to compliance. It comes back with the same notes as last time. The Compliance Flagger reads your draft first - so it arrives at sign-off already argued with once.
Every draft that bounces off compliance costs a review cycle: the performance figure with no balance, the "you should" that reads like advice, the risk warning three paragraphs from the claim it belongs to. Redraft, resubmit, wait again - and the content calendar quietly slips.
Six of the patterns sign-off most often stops content for.
Each flag: the exact quote, a severity (flag / consider), the reason in plain English, and a rewording where it's safe to offer one. Where the fix needs your firm's approved wording - risk warnings, disclosures - it marks the spot instead of inventing one.
Buy it once. Use it in the Claude you already have.
Buy once, download the skill file.
One payment, then download from your account - a magic-link sign-in, no password. No subscription.
Drop it into your Claude.
Upload it to claude.ai as a skill. It sits alongside how you already work.
Paste a draft and ask for a check.
It flags the passages sign-off is likely to stop, with reasons - before the draft leaves your desk.
Yours to keep. No subscription - just a magic-link sign-in to collect it.
UK advisers, planners, and accountancy firms who write their own content.
Newsletters, web pages, social posts, client letters - anything that goes through a sign-off that keeps catching the same things. If your content goes out with no sign-off process at all, this skill will make you a more careful writer. It will not stand in for the process you should have.
- Not a compliance service - it produces no compliance conclusions and never says content is "compliant" or "approved". Nothing that runs in a chat window honestly can.
- An assist to your sign-off, not a replacement. Every output ends by saying exactly that.
- No rule numbers as verdicts, no regulatory advice. It flags the patterns experienced sign-off catches, in plain English, and leaves the judgement with you and your compliance support.
- Language patterns only. It doesn't check facts, figures, statutory status lines, prescribed warnings, or who your content is aimed at - your sign-off checks all of those. A clean run means clean of its six categories, not clean of everything.
The file is markdown. The judgement is the product.
The paid skills market is full of cheap files that are - as one developer who audited the market put it - "a name, a vague description", and no change in behaviour. This one is built on the judgement sign-off reviewers apply: six of the patterns they most often stop content for, the register a regulated firm has to hold, and worked examples that show the behaviour before you buy.
Want the version built on your firm's own voice, approved wordings and document types? That's the regulated pack - a done-for-you engagement, by conversation. See the regulated line. This skill is the self-serve starting point.
Clear first. In your voice. Then flagged.
Plain English for the clarity half of fair communication; Voice Match so the fixes still sound like you.
The questions a careful buyer should ask.
Does this make my content compliant?
No, and be wary of anything that says it can. It prepares content for the person who signs it off - fewer cycles, calmer reviews. The judgement stays human.
Is this advice?
No. The skill reviews the language of draft content. It doesn't know your clients, your permissions, or your firm's approved wordings, and it doesn't pretend to.
Will it work for my paraplanner or marketing person?
Yes - it's most useful to whoever writes the first draft. The team licence covers up to five people in one organisation.
What happens to the text I paste in?
The skill reviews draft marketing content - newsletters, posts, pages - not client data, and client-specific documents shouldn't go through it. The draft is processed in your own claude.ai account under Anthropic's data terms; point your firm's data-protection review at those terms, the same as for any Claude use.
I'm not UK-based - is it still useful?
It's calibrated for UK client-facing content. Elsewhere, treat the flags as prompts rather than calibration.
Arrive at sign-off already argued with once.
Flags what your compliance sign-off is likely to catch - the exact quote, the reason in plain English, and a rewording where it's safe to offer one.
Personal licence for your own use, or a Team licence for up to 5 people in one organisation. Not for resale or redistribution. An assist to your sign-off process, not a compliance service.