amazon-product-research-agent-skills

Amazon Supplier Feasibility

Purpose and scope

Turn a product specification into a supplier-ready feasibility package and an auditable comparison. This Skill prepares questions and evaluates supplied evidence; it does not contact factories, negotiate, order samples, or approve production unless separately authorized.

Implementation Basis and Dependencies

The method combines specification completeness, RFQ normalization, process-risk review, cost and lead-time comparison, and predeclared sample gates.

Dependency	Required	Validation	Fallback
Product specification	Yes	Materials, dimensions, performance, packaging, and target volume are stated or marked TBD	Return specification gaps before RFQ
Supplier quotes	Conditional for comparison	Quote date, currency, Incoterm, MOQ, tooling, lead time, and inclusions are present	Produce RFQ only
Compliance requirements	Conditional	Marketplace and product category are defined	Flag compliance as unresolved
Supplier communication access	Optional	User separately authorizes contact and account use	Prepare messages without sending
Sample and inspection capability	Optional	Test method and responsible party are identified	Produce gates without claiming approval

Inputs

Input	Requirement	Format and handling
Product specification	Required	Structured document, drawing, image, BOM, or written requirements with units and tolerances
Demand assumptions	Required for MOQ planning	Initial order quantity, monthly demand range, target launch date, and replenishment window
Commercial targets	Conditional	Target EXW/FOB/landed cost, currency, Incoterm, tooling budget, payment terms, and cash limit
Supplier quotes	Optional	PDF, XLSX, CSV, email text, or form response; preserve source and quote date
Quality and compliance needs	Conditional	Critical-to-quality dimensions, performance tests, labels, packaging, certifications, and marketplace

Supplier identities, contacts, pricing, and agreements may be confidential. Keep them local unless external processing is authorized.

Outputs

1. RFQ package with product scope, forecast tiers, commercial questions, quality requirements, and required evidence.
2. Normalized supplier comparison separating unit cost, tooling, packaging, freight basis, MOQ, lead time, terms, and exclusions.
3. Manufacturing-risk register with likelihood, impact, evidence, mitigation, and owner.
4. Sample acceptance plan with test method, sample size, threshold, failure action, and required records.
5. Feasibility status: READY_FOR_RFQ, READY_FOR_SAMPLE, CONDITIONAL, or INSUFFICIENT_EVIDENCE.

Success requires comparable commercial bases and no hidden assumptions in cost or lead time. The Skill does not send messages or place orders by default.

Workflow

1. Check the specification for missing materials, dimensions, tolerances, performance, packaging, and regulatory requirements.
2. Separate non-negotiable requirements from target and stretch requirements.
3. Build an RFQ using the bundled assets/supplier-quote-template.csv fields.
4. Normalize all quotes to the same quantity tier, currency, Incoterm, inclusions, and evidence date.
5. Identify process, tooling, capacity, component, quality, compliance, packaging, and logistics risks.
6. Define sample tests and acceptance thresholds before samples arrive.
7. Compare suppliers on total feasibility, not unit price alone.
8. Route cost scenarios to amazon-unit-economics-cashflow before recommending an order.

Example

For a magnetic phone mount, the RFQ should distinguish magnet grade, adhesive system, ball-joint torque, coating, temperature range, phone-weight test, packaging, tooling ownership, replacement parts, and quote basis. A low unit quote without those fields is not comparable.

Error handling and stopping conditions

• Missing specification: stop before supplier ranking.
• Quotes use different Incoterms or inclusions: normalize or mark incomparable.
• Certification claims lack documents: treat as unverified.
• Supplier contact or ordering is requested without authorization: prepare the action and stop before sending.
• Sample test fails a critical gate: do not average it away with price advantages.

Validation and definition of done

• Every quote includes or explicitly lacks each normalized field.
• Critical manufacturing and quality risks have evidence and an owner.
• Sample gates are measurable and declared before results.
• Feasibility status names the remaining blocker and next action.

This site is open source. Improve this page.