# Price Client AI Feedback

> Sorts a client's chatbot review of your work against the agreed brief into points that are right, points that contradict the brief, points out of scope and points that are wrong, then drafts a calm reply and shows what the next round costs. Use for "run price-client-ai-feedback", "the client ran my work through a chatbot", "they say AI found problems in my report", "the client says I am doing a bad job", "triage this feedback", "is this feedback in scope", "how many revision rounds have we used", part of the Pricing Under AI Pressure Pack by 90n.

- Skill: `polar-bear-org/price-client-ai-feedback` (Agent Skill)
- Install (CLI): `npx get-skill add polar-bear-org/price-client-ai-feedback`
- Raw SKILL.md: https://api.skillmd.com/api/skills/polar-bear-org/price-client-ai-feedback/raw
- Safety review: pending
- Works with: Claude Code, Claude.ai, OpenAI Codex
- Category: Docs & Writing
- Author: polar-bear-org (https://skillmd.com/u/polar-bear-org)
- Updated: 2026-11-14
- Page: https://skillmd.com/skills/polar-bear-org/price-client-ai-feedback

---


# Client AI Feedback Triage

## When To Use
The client ran your work through a chatbot and now says you are doing a bad job, with a long list that mixes fair points, new requests and things that are simply wrong. This answers one question: which points do you fix inside the agreed rounds, which contradict what was agreed, and which are new work with a price?

## When Not To Use
If one point is a clear new request that needs pricing in detail, run Scope Change Request. If the work has not started and the problem is the brief itself, run AI-Written Brief Review.

## Inputs
- The feedback as the client sent it, chatbot output included
- The agreed brief or Statement of Work, with acceptance criteria and the revision rounds agreed
- The delivered work, and your rates for extra rounds or changes
If you have no agreed brief, I sort against what you can show was agreed (emails, a proposal) and mark the output as a first draft.

## Approach
Triage against the agreed brief and acceptance criteria, a practitioner convention: every point is judged against what was agreed, not against what a chatbot would prefer. A model reviewing work without the brief will happily ask for a different piece of work. The failure this prevents is the unpaid rewrite: you fix everything to prove you are good, and the next round arrives the same way.

## Workflow
1. Ask, all at once: where is the agreed brief or Statement of Work? How many revision rounds were agreed, and how many are used? Is there a deadline on your reply?
2. Split the feedback into single points and number them, keeping the client's wording. Merge only exact repeats.
3. Put each point in one bin. Right: a fair point inside the brief, fixed inside the agreed rounds. Contradicts the brief: quote the brief line it conflicts with. Out of scope: new work, routed to a change request. Wrong: state the evidence from the work (page, figure, source).
4. Be fair in both directions. A good point from a chatbot is still a good point; put it in "right" and fix it. A point you dislike is not "wrong" without evidence.
5. Count rounds used against rounds agreed. If this round goes beyond them, the next round's cost comes from your rates, or a [your figure] placeholder.
6. Draft a calm reply: thanks, what you will fix and by when, where a point conflicts with the agreed brief (quoting it), what is a change and its price, an offer to talk it through. Neutral about the client's chatbot use. You edit and send it.

## Output Format
```markdown
# Client AI Feedback Triage
Client: [name or code] · Work: [deliverable] · Feedback received: [date]
## Points
| # | Point (client's wording) | Bin | Basis |
|---|---|---|---|
| [1] | [text] | [right / contradicts brief / out of scope / wrong] | [brief line / evidence / change] |
## Summary
Right: [count] · Contradicts brief: [count] · Out of scope: [count] · Wrong: [count]
## Rounds
Agreed: [number] · Used including this one: [number] · Next round costs: [your figure]
## Changes to price
| Point | Change | Price |
|---|---|---|
| [#] | [what] | [your figure] |
## Reply draft
[Calm reply in your voice]
## Decision
[You] decide by [date] which fixes go into this round and whether the changes are quoted, and you send the reply yourself.
```

## Done When
- Every point sits in exactly one bin with its basis
- Each "contradicts brief" point quotes the brief line
- Each "wrong" point cites evidence from the work
- Rounds used are counted against rounds agreed

## Quality Bar
- The feedback is triaged, never the client contact
- No mockery of the chatbot or of the client for using it
- Prices for changes and rounds come from your rates only
- Nothing is added to the feedback that the client did not send
- The person sends the reply; Claude drafts it

## Next
Run price-ai-disclosure-policy (AI Disclosure Policy) to set clear rules on AI in your own work.

## About the makers

This pack is made by 90n ([90n.ai](https://90n.ai)): AI-native ways of working and tools for Product, Marketing and HR teams, built by ex-McKinsey founders. We map where time and money leak, redesign the work around AI and stay until your people run it themselves, and we run our own company on Claude. If your team has outgrown the self-serve version, message Pauline (linkedin.com/in/paulinebertry).

