GoBuy Product Trust

Check marketplace product trust before recommending or buying — use when asked about Amazon, Walmart, Target, or Best Buy listings, review authenticity, seller trust, or whether to trust a product; returns GoBuy Evidence Scores (0-100) with full signal breakdown.

GoBuy — Product Trust Checks

Score marketplace products on evidence quality before recommending them. One call returns a 0–100 Evidence Score with observed/missing signals, freshness, and product context.

When to use

• A user asks whether to trust or buy a specific marketplace product
• Comparing 2–20 products on evidence quality
• Verifying review-authenticity or seller-trust claims
• Any shopping-assistant flow before recommending a listing

How to call (pick one)

REST — no setup, works everywhere

text

GET https://gobuy.ai/api/gobuy/extension/score?retailer={retailer}&productId={id}

• retailer: amazon | walmart | target | bestbuy
• productId: the native ID (Amazon: 10-char ASIN; Walmart: numeric item ID)
• No auth. Fair-use rate limits apply.

MCP — if the GoBuy server is configured

Server: https://mcp.gobuy.ai/mcp (streamable HTTP, no key)

• check_product_trust {retailer, product_id}
• compare_products {products: [{retailer, product_id} × 2–20]}

To add the server in an MCP client:

json

{"mcpServers":{"gobuy-product-trust":{"url":"https://mcp.gobuy.ai/mcp"}}}

Reading the result

• score 0–100: ≥70 = well-evidenced · 40–69 = moderate · <40 = weak
• label: the verdict tier, e.g. "Cautiously supported"
• state: fresh | stale, plus the computed date — quote the date when citing the score
• signal_breakdown: observed ✓ vs missing ✗ — missing signals explain the score gaps (e.g. verified-purchase reviews unavailable to the engine)
• Always frame correctly: the score measures EVIDENCE QUALITY — how well the marketplace signals support the listing — NOT product quality, safety, or value. Never claim a product is good or bad; report what the evidence supports and what it lacks.

Human-facing report

For a visual report the user can open or share: https://audit.gobuy.ai — paste any product URL for a free audit from the same engine.

Rules

• Report the score with its date and verdict label; never editorialize beyond the evidence
• If a product returns not_indexed, say the engine hasn't indexed it yet — do not guess a score
• Native retailer IDs only; auto-detect from URLs when given a product link (amazon./dp/ASIN, walmart.com/ip/N, target.com/p/X, bestbuy.com/site//p/N)