shopping-aggregatorlisted

# shopping-aggregator A thin orchestration layer for consumer shopping price comparison. It does **only three things nothing else does**: (1) parse a buy intent into a triage of the right shopping domains + channel classes; (2) detect which specialized shopping MCP / extension / OSS tool is connected and guide installing missing ones; (3) enforce **price-specific** guardrails (snapshot timestamp, stock state, landed-cost normalization, coupon-cart verification, retailer trust tiers). The heavy lifting, live page fetch, history backfill, adversarial verification, citation synthesis, is **delegated** to playwright / BigGo / Keepa MCPs / `deep-research` / `market-intel`. Do not re-implement those. > **Design philosophy (governs all changes): root-cause design, not incremental patching**, fix the > assumption underneath a problem, not the symptom on top. Full statement in `PHILOSOPHY.md`; every > change must pass "does it fix the framing, or just patch a symptom?" ## Scope, when this skill applies, and when to stop The frontmatter `description` carries **trigger phrases only**. Which retailers are covered (Amazon / eBay / Walmart / Best Buy / Taobao / JD / PDD, see the Region row in Step 1), which price-history and coupon layers are used (Keepa / Camelcamelcamel / 慢慢买, coupon-cart verification), and how lodging booking is driven all live in this file. Keep them here: `description` is charged against the library metadata budget on every turn, this file loads only when the skil