Canonical page: https://www.kuudo.com/features/skills/amazon-seller/amazon-repricer/

Amazon Seller Skills v0.2.1 Reference

# Run an Amazon Repricer From Computed Floors

Kuudo's Amazon repricer runs in your AI client. It reads current prices, competing offers, fees and stock through the Selling Partner MCP, then computes a floor for each SKU and proposes price changes for approval. Choose this workflow when you want to inspect the costs and rules behind each price. The shipped Skill validates proposed changes with Amazon and requires human approval. You can edit the Skill to allow unattended application under your own rules.

Proposes; writes on approval

[Try the starter prompt](#starter-prompt)[Connect Skills](/docs/quick-start/chatgpt-skills/)

Example output visualization: Change-set review

Change-set review

skill output

\-Match the lowest competitor, keep 20% margin across the catalog

+Floor per SKU at the candidate price, comparable offers only, cooldown between moves

Floors computed from fee estimates

Validation preview reported

Change-set awaiting approval

Primary input

Listings, offers, fee estimates, and stock from SP-API, or your own figures in advisory mode

Primary output

Per-SKU floors and ceilings

[

Powered by

Amazon Selling Partner MCP

](/features/amazon-selling-partner-mcp/)[

Source contract

amazon-sp-repricing

PolyForm-Shield-1.0.0

](https://github.com/KuudoAI/skills/blob/674e68adc1c0242e5ea440583f82d26c93c86b1d/skills/amazon-sp-repricing/SKILL.md)

Is this the right Skill?

## Use it for a clear operating job.

Best for

*   Losing the Buy Box on part of the catalog
*   Floors and ceilings for a repricer or Automate Pricing
*   Breaking a price war with one competitor

Use another path when

Advertising bids or budgets, listing copy or images, inventory forecasting, or returns analysis.

Starter prompt

## Start with the task, then add your scope.

Copy this into a connected client and replace the account, date, ASIN, or threshold details with yours.

```text
I'm losing the Buy Box on 30% of my listings. I have 200 SKUs in electronics, price range $15 to $80. Pull the data, compute a floor per SKU, segment the catalog, and give me a change-set to approve, not applied prices.
```

The Skill will request any required context or approval before guarded actions.

The outcome

## A defined result, not a generic answer.

A repricing decision document with per-SKU floors, segments and rules, guardrails, and a change-set where every proposed price sits beside its floor and the rule that produced it.

Inputs

*   Listings, offers, fee estimates, and stock from SP-API, or your own figures in advisory mode
*   Landed cost, minimum contribution, storage, return rate
*   SKUs excluded for promotional or contractual reasons

Outputs

*   Per-SKU floors and ceilings
*   Segments, rule archetypes, and guardrails
*   Validated change-set with each price beside its floor

Workflow

## The playbook your agent follows.

1.  01
    
    ### Pull the data
    
    Take current prices, competing offers at landed price, featured-offer standing, fees, and stock cover from the API, then ask the seller only for what it cannot supply.
    
2.  02
    
    ### Compute floors
    
    Evaluate fees at the candidate price per SKU, add storage, returns, and advertising allocation, and set a ceiling too.
    
3.  03
    
    ### Diagnose the Featured Offer
    
    Separate eligibility, landed price, fulfillment, and nobody-winning cases, since lowest price does not decide it.
    
4.  04
    
    ### Segment and write rules
    
    Split the catalog on competitive structure and margin headroom, then assign hold, match, beat by increment, target expected price, seek high, or stock-out defence.
    
5.  05
    
    ### Validate and propose
    
    Run every change through the listings validation preview, report what it returns, and hand back the change-set for approval.
    
6.  06
    
    ### Measure the pair
    
    Report featured-offer share beside contribution per unit, so a bought offer cannot look like a win.
    

Guardrails

## What keeps the workflow dependable.

The shipped Skill proposes a change-set for human approval; unattended application requires you to modify the Skill and set your own rules

Every proposed listing price runs through patchListingsItem with mode=VALIDATION\_PREVIEW before any write

It can also drive Amazon's automated pricing rules, which hand ongoing price moves to Amazon; a rule change has no validation preview, is shown as unvalidated, and waits for human approval

Featured-offer share is reported beside contribution per unit, not on its own

Connected systems

## The context and tools it coordinates.

Amazon Selling Partner MCP

Product Pricing and Fees APIs

Listings Items validation preview

Keep building

## Related workflows and field guides.

[All Amazon Seller Skills →](/features/skills/amazon-seller/)

[Related SkillSales and Traffic analyzer](/features/skills/amazon-seller/sales-traffic-analyzer/)[Related SkillBudget pacing](/features/skills/amazon-ads/budget-pacing/)[Product surfaceAmazon Selling Partner MCP](/features/amazon-selling-partner-mcp/)

amazon-sp-repricing

## Give your agent the complete playbook.

Connect Skills once, then start with the prompt above whenever this job comes up.

[View prompt](#starter-prompt)[Connect Skills](/docs/quick-start/chatgpt-skills/)