See the churn coming while you can still do something about it.
The Tool that reads your client book for the relationships about to go quiet, diagnoses why, gives you a specific intervention for each, and tells you which clients to fight for and which to release.
Clients rarely leave suddenly. They disengage first, in a sequence of small signals that are obvious in hindsight and invisible in the moment because each one alone is explainable. The meeting that gets shorter. The decision-maker who stops joining the call. The brief that gets vaguer. By the time the relationship is openly in trouble, the decision to leave was usually made weeks earlier. Client Risk reads the book for those signals while there's still time to act. It rates each relationship, diagnoses the actual cause, a lost champion, a plateau, budget pressure, a relationship that was always transactional, because the intervention depends entirely on the cause, then writes a specific intervention matched to that cause rather than a vague resolution to check in. And it says the thing most owners avoid: some clients should be fired, and keeping them costs the capacity to win a better one. Your client roster is commercially sensitive. The Tool only needs your own honest read on each relationship, never the clients' confidential data, and you should keep both the input and the output somewhere you control.
What it actually does.
Eight retained clients, and a feeling that two are wobbling. You paste the roster with honest context. Client Risk rates them: your largest client is HIGH, the senior sponsor has dropped off the last three calls and sends a junior now, the briefs are vaguer. The cause is a lost champion, the person who bought the relationship has disengaged. The intervention isn't a check-in, it's a deliberate rebuild with the new senior contact, a 30-minute session framed around their priorities, and the signal of success is the sponsor re-engaging within two sessions. A small, low-margin client you dread is also HIGH, but the diagnosis is mutual disengagement, so that one is a managed exit, not a rescue. The Tool ranks by value and savability: fight for the largest this week, hold the steady ones, release the one that's draining you. Next quarter you re-read it to see whether the rebuild held and whether releasing the drainer freed the capacity it was meant to.
- ✦Agency and consultancy owners with a book of ongoing clients
- ✦Operators running quarterly account reviews
- ✦Anyone who's lost a client to slow disengagement they noticed too late
- →client-risk.zip — a native Claude Skill. Upload to Settings → Skills in Claude, or unzip to
~/.claude/skills/ - →client-risk-skill.md — plain markdown. Works in ChatGPT Custom GPTs, Gemini, Claude Projects, agent system prompts — anywhere
- →A README inside the zip with install steps for Claude, ChatGPT, Gemini, and the API
No subscription · free updates within v1.x
Works as a Claude Skill (.zip) and as a plain .md for ChatGPT, Gemini & every LLM
Delivered to your email within 60 seconds · Secured by Stripe
Price includes UK VAT where applicable
Before you buy.
How is this different from just asking Claude directly?
Ask Claude about a wobbling client and it suggests a check-in. Client Risk rates the whole book, diagnoses the actual cause of each at-risk relationship (a lost champion, a plateau, budget pressure, an always-transactional account), writes an intervention matched to that specific cause, and ranks by value and savability — including telling you which clients to release. It says the part most owners avoid: some clients cost you the capacity to win a better one.
Will this work with ChatGPT, Gemini, and other LLMs?
Yes. Native Claude Skill (.zip) plus a plain .md for ChatGPT, Gemini, or any LLM. You paste your own honest read of each relationship and the procedure runs anywhere.
Can I share this with my team?
It's licensed for your own use; the risk read it produces is yours to use in account reviews. Note the input is commercially sensitive — keep both the roster you paste and the output somewhere you control. For team access, email hello@authority.md about team licensing.
What if it doesn't work for me?
If the file hasn't been delivered, or is defective or empty, we refund in full — see the refund policy in our terms. It works from your own assessment of each relationship, never the clients' confidential data — it's a judgement aid, not a substitute for actually talking to the client.