{"id":"e4bfe17e-92f3-4882-836f-1611e84ccbda","entityType":"agent","slug":"clawhub-harrylabsj-taobao-competitor-analyzer","name":"Taobao Price Compare","canonicalUrl":"https://www.xpersona.co/agent/clawhub-harrylabsj-taobao-competitor-analyzer","canonicalPath":"/agent/clawhub-harrylabsj-taobao-competitor-analyzer","generatedAt":"2026-10-09T18:36:30.859Z","source":"CLAWHUB","claimStatus":"UNCLAIMED","verificationTier":"NONE","summary":{"evidence":{"source":"CLAWHUB","verified":false,"confidence":"medium","updatedAt":"2026-10-09T16:47:14.132Z","emptyReason":null},"description":"Taobao price and competitor comparison assistant. Input a Taobao title/link or product need; compare JD, PDD, and Vipshop using browser-visible evidence, nor...","descriptionLabel":"Source description","evidenceSummary":"Capability contract not published. No trust telemetry is available yet. 2.3K downloads reported by the source. Last updated 10/9/2026.","installCommand":"clawhub skill install s17a8m9q4jybb46cv60h4fxard83hmsn:taobao-competitor-analyzer","sourceUrl":"https://clawhub.ai/harrylabsj/taobao-competitor-analyzer","homepage":"https://clawhub.ai/harrylabsj/skills/taobao-competitor-analyzer","primaryLinks":[{"label":"View on ClawHub","url":"https://clawhub.ai/harrylabsj/taobao-competitor-analyzer","kind":"source"},{"label":"Homepage","url":"https://clawhub.ai/harrylabsj/skills/taobao-competitor-analyzer","kind":"homepage"}],"safetyScore":84,"overallRank":62,"popularityScore":59,"trustScore":null,"claimedByName":null,"isOwner":false,"seoDescription":"Taobao Price Compare technical dossier on Xpersona with agent coverage, OPENCLEW support, and live trust metadata."},"coverage":{"evidence":{"source":"public-profile","verified":false,"confidence":"medium","updatedAt":"2026-10-09T16:47:14.132Z","emptyReason":null},"protocols":[{"protocol":"OPENCLEW","label":"OpenClaw","status":"self-declared","notes":"Declared in the public agent profile."}],"capabilities":[],"verifiedCount":0,"selfDeclaredCount":1,"capabilityMatrix":{"rows":[{"key":"OPENCLEW","type":"protocol","support":"unknown","confidenceSource":"profile","notes":"Listed on profile"}],"flattenedTokens":"protocol:OPENCLEW|unknown|profile"}},"adoption":{"evidence":{"source":"CLAWHUB","verified":false,"confidence":"medium","updatedAt":"2026-10-09T16:47:14.132Z","emptyReason":null},"stars":null,"forks":null,"downloads":2275,"packageName":null,"latestVersion":"2.1.1","tractionLabel":"2.3K downloads"},"release":{"evidence":{"source":"CLAWHUB","verified":false,"confidence":"medium","updatedAt":"2026-10-09T16:47:14.132Z","emptyReason":null},"lastUpdatedAt":"2026-10-09T16:47:14.132Z","lastCrawledAt":"2026-10-09T16:47:14.132Z","lastIndexedAt":null,"nextCrawlAt":"2026-10-10T16:47:14.132Z","lastVerifiedAt":null,"highlights":[{"version":"2.1.1","createdAt":"2026-06-23T11:33:25.857Z","changelog":"Add Taobao baseline bridge, evidence gate, and expanded proxy evals.","fileCount":6,"zipByteSize":19149},{"version":"2.1.0","createdAt":"2026-06-17T02:36:27.793Z","changelog":"P1: reposition as Taobao price comparison and standardize shopping safety boundaries.","fileCount":5,"zipByteSize":16529},{"version":"1.2.0","createdAt":"2026-06-07T13:28:22.529Z","changelog":"Add category-aware buying playbooks, preference routing, platform-fit gates, category-specific price thresholds, and proxy eval cases.","fileCount":6,"zipByteSize":18727},{"version":"1.1.0","createdAt":"2026-06-07T12:49:13.691Z","changelog":"Add Taobao baseline handling, 100-point match scoring, normalized price comparison, evidence gates, and recommendation strength.","fileCount":5,"zipByteSize":12821},{"version":"1.0.3","createdAt":"2026-03-31T10:00:34.737Z","changelog":"Add matrix cross-sell guidance to route users to the right shopping skill.","fileCount":4,"zipByteSize":7415},{"version":"1.0.2","createdAt":"2026-03-31T09:42:52.816Z","changelog":"Improve install-page positioning, add stronger trigger language, and clarify shopping decision output.","fileCount":4,"zipByteSize":7200},{"version":"1.0.1","createdAt":"2026-03-31T09:39:57.474Z","changelog":"Upgrade from price table output to buying recommendation and seller-risk comparison workflow.","fileCount":3,"zipByteSize":6013},{"version":"1.0.0","createdAt":"2026-03-14T11:55:52.660Z","changelog":"Initial release with browser-only competitor price analysis across JD, Pinduoduo, and Vipshop.","fileCount":3,"zipByteSize":4184}]},"execution":{"evidence":{"source":"CLAWHUB","verified":false,"confidence":"low","updatedAt":null,"emptyReason":"No published capability contract is available yet."},"installCommand":"clawhub skill install s17a8m9q4jybb46cv60h4fxard83hmsn:taobao-competitor-analyzer","setupComplexity":"low","setupSteps":["Install using `clawhub skill install s17a8m9q4jybb46cv60h4fxard83hmsn:taobao-competitor-analyzer` in an isolated environment before connecting it to live workloads.","No published capability contract is available yet, so validate auth and request/response behavior manually.","Review the upstream CLAWHUB listing at https://clawhub.ai/harrylabsj/taobao-competitor-analyzer before using production credentials."],"contract":{"contractStatus":"missing","authModes":[],"requires":[],"forbidden":[],"supportsMcp":false,"supportsA2a":false,"supportsStreaming":false,"inputSchemaRef":null,"outputSchemaRef":null,"dataRegion":null,"contractUpdatedAt":null,"sourceUpdatedAt":null,"freshnessSeconds":null},"invocationGuide":{"preferredApi":{"snapshotUrl":"https://www.xpersona.co/api/v1/agents/clawhub-harrylabsj-taobao-competitor-analyzer/snapshot","contractUrl":"https://www.xpersona.co/api/v1/agents/clawhub-harrylabsj-taobao-competitor-analyzer/contract","trustUrl":"https://www.xpersona.co/api/v1/agents/clawhub-harrylabsj-taobao-competitor-analyzer/trust"},"curlExamples":["curl -s \"https://www.xpersona.co/api/v1/agents/clawhub-harrylabsj-taobao-competitor-analyzer/snapshot\"","curl -s \"https://www.xpersona.co/api/v1/agents/clawhub-harrylabsj-taobao-competitor-analyzer/contract\"","curl -s \"https://www.xpersona.co/api/v1/agents/clawhub-harrylabsj-taobao-competitor-analyzer/trust\""],"jsonRequestTemplate":{"query":"summarize this repo","constraints":{"maxLatencyMs":2000,"protocolPreference":["OPENCLEW"]}},"jsonResponseTemplate":{"ok":true,"result":{"summary":"...","confidence":0.9},"meta":{"source":"CLAWHUB","generatedAt":"2026-10-09T18:36:30.854Z"}},"retryPolicy":{"maxAttempts":3,"backoffMs":[500,1500,3500],"retryableConditions":["HTTP_429","HTTP_503","NETWORK_TIMEOUT"]}},"endpoints":{"dossierUrl":"https://www.xpersona.co/api/v1/agents/clawhub-harrylabsj-taobao-competitor-analyzer/dossier","snapshotUrl":"https://www.xpersona.co/api/v1/agents/clawhub-harrylabsj-taobao-competitor-analyzer/snapshot","contractUrl":"https://www.xpersona.co/api/v1/agents/clawhub-harrylabsj-taobao-competitor-analyzer/contract","trustUrl":"https://www.xpersona.co/api/v1/agents/clawhub-harrylabsj-taobao-competitor-analyzer/trust"}},"reliability":{"evidence":{"source":"runtime-metrics","verified":false,"confidence":"low","updatedAt":null,"emptyReason":"No trust, reliability, or runtime telemetry is available."},"trust":{"status":"unavailable","handshakeStatus":"UNKNOWN","verificationFreshnessHours":null,"reputationScore":null,"p95LatencyMs":null,"successRate30d":null,"fallbackRate":null,"attempts30d":null,"trustUpdatedAt":null,"trustConfidence":"unknown","sourceUpdatedAt":null,"freshnessSeconds":null},"decisionGuardrails":{"doNotUseIf":["Contract metadata is missing or unavailable for deterministic execution."],"safeUseWhen":[],"riskFlags":["missing_or_unavailable_contract","trust_data_unavailable","schema_references_missing"],"operationalConfidence":"low"},"executionMetrics":{"observedLatencyMsP50":null,"observedLatencyMsP95":null,"estimatedCostUsd":null,"uptime30d":null,"rateLimitRpm":null,"rateLimitBurst":null,"lastVerifiedAt":null,"verificationSource":null},"runtimeMetrics":{"successRate":null,"avgLatencyMs":null,"avgCostUsd":null,"hallucinationRate":null,"retryRate":null,"disputeRate":null,"p50Latency":null,"p95Latency":null,"lastUpdated":null}},"benchmarks":{"evidence":{"source":"no-benchmark-data","verified":false,"confidence":"low","updatedAt":null,"emptyReason":"No benchmark suites or observed failure patterns are available."},"suites":[],"failurePatterns":[]},"artifacts":{"evidence":{"source":"CLAWHUB","verified":false,"confidence":"medium","updatedAt":"2026-10-09T16:47:14.132Z","emptyReason":null},"readme":"Skill: Taobao Price Compare\n\nOwner: harrylabsj\n\nSummary: Taobao price and competitor comparison assistant. Input a Taobao title/link or product need; compare JD, PDD, and Vipshop using browser-visible evidence, nor...\n\nTags: buying-advice:2.1.1, category-aware:2.1.1, china:2.1.1, comparison:2.1.1, competitor-analysis:2.1.1, ecommerce:2.1.1, latest:2.1.1, price-analysis:1.2.0, price-compare:2.1.1, shopping:2.1.1, shopping-decision:2.1.1, taobao:2.1.1\n\nVersion history:\n\nv2.1.1 | 2026-06-23T11:33:25.857Z | user\n\nAdd Taobao baseline bridge, evidence gate, and expanded proxy evals.\n\nv2.1.0 | 2026-06-17T02:36:27.793Z | user\n\nP1: reposition as Taobao price comparison and standardize shopping safety boundaries.\n\nv1.2.0 | 2026-06-07T13:28:22.529Z | user\n\nAdd category-aware buying playbooks, preference routing, platform-fit gates, category-specific price thresholds, and proxy eval cases.\n\nv1.1.0 | 2026-06-07T12:49:13.691Z | user\n\nAdd Taobao baseline handling, 100-point match scoring, normalized price comparison, evidence gates, and recommendation strength.\n\nv1.0.3 | 2026-03-31T10:00:34.737Z | user\n\nAdd matrix cross-sell guidance to route users to the right shopping skill.\n\nv1.0.2 | 2026-03-31T09:42:52.816Z | user\n\nImprove install-page positioning, add stronger trigger language, and clarify shopping decision output.\n\nv1.0.1 | 2026-03-31T09:39:57.474Z | user\n\nUpgrade from price table output to buying recommendation and seller-risk comparison workflow.\n\nv1.0.0 | 2026-03-14T11:55:52.660Z | user\n\nInitial release with browser-only competitor price analysis across JD, Pinduoduo, and Vipshop.\n\nArchive index:\n\nArchive v2.1.1: 6 files, 19149 bytes\n\nFiles: README.md (3911b), references/category-playbooks.md (7975b), references/site-notes.md (6507b), skill-card.md (2536b), SKILL.md (21329b), _meta.json (145b)\n\nFile v2.1.1:SKILL.md\n\n---\nname: taobao-competitor-analyzer\ndescription: \"Taobao price and competitor comparison assistant. Input a Taobao title/link or product need; compare JD, PDD, and Vipshop using browser-visible evidence, normalize prices, score same-item confidence, and output where to buy. Safe boundary: no login, no order submission, no payment.\"\n---\n# Taobao Competitor Analyzer\n\nCompare a Taobao product with the same or closest-matching listings on 京东、拼多多、唯品会 using the browser tool only. Work from a product name, Taobao title, link, screenshot, or user-provided baseline details. Return a compact purchase decision with visible price evidence, normalized comparability, match score, category-aware platform fit, recommendation strength, and risk notes.\n\nWhat makes this skill useful:\n- It compares comparable items instead of chasing misleading low prices.\n- It separates Taobao baseline price from competitor prices.\n- It normalizes price by spec, quantity, shipping, and promo conditions before ranking.\n- It scores same-item confidence before making a buying recommendation.\n- It applies category-specific risk thresholds instead of treating shoes, phones, skincare, and paper towels the same way.\n- It explains whether a lower price depends on coupons, membership, subsidy, or group-buy.\n- It ends with a clear buy / wait / avoid recommendation instead of just a table.\n\n## When To Use\n\nUse this skill when the user is effectively asking:\n- 这件淘宝商品别的平台多少钱\n- 有没有同款或更划算的平台\n- 京东 / 拼多多 / 唯品会 哪个更值得买\n- 这几个平台价格差这么多正常吗\n- 帮我做一个同款比价和购买建议\n- 这个淘宝商品换平台买值不值\n- 同样价差下哪个平台风险更低\n\nThe skill should optimize for purchase decisions, not raw data collection.\n\n## Commerce Matrix\n\nThis skill is the cross-platform comparison node in the shopping matrix.\n\nPrefer nearby skills when the task is narrower:\n- `taobao-shopping` for Taobao-only listing and seller evaluation\n- `jd-shopping` for trust-first self-operated buying\n- `pdd-shopping` for low-price and subsidy-first buying\n- `tianmao` for flagship-store and authenticity-first buying\n- `vip` for branded discount and flash-sale buying\n- `alibaba-shopping` when the user first needs to choose between Taobao, Tmall, and 1688\n\n## Taobao Baseline Bridge\n\nUse `taobao-shopping` first, or apply its same discipline yourself, when the Taobao side is not yet clear.\n\nBefore cross-platform comparison, establish the Taobao baseline:\n\n- exact product identity\n- selected SKU attributes\n- visible or user-provided Taobao price\n- store type and seller trust cue\n- coupon / membership / threshold condition\n- return, warranty, invoice, authenticity, or delivery caveat when relevant\n\nIf the user only provides a Taobao title with no price or listing evidence, this skill can compare competitor visible offers, but it must not claim the user should switch away from Taobao. Write `淘宝基准价未提供`.\n\nIf the user's real question is only \"is this Taobao listing trustworthy?\", stay in or route to `taobao-shopping` instead of broadening into JD/PDD/Vipshop.\n\n## Workflow\n\n1. Normalize the Taobao baseline.\n   - Extract product identity: brand, model / series, variant, size / spec / count, color / flavor / version, packaging, and must-match attributes.\n   - Record the Taobao price only if the user provides it or it is browser-visible.\n   - If no Taobao price is available, write `淘宝基准价未提供` and do not imply whether switching platforms saves money versus Taobao.\n2. Classify the product category and buying posture.\n   - Pick the closest category from `references/category-playbooks.md`.\n   - Infer whether the user is price-first, trust-first, speed-first, exact-variant-first, or research-first.\n   - If the category or posture would change the recommendation and cannot be inferred, ask one short follow-up.\n3. If the input is only a Taobao-style long title, compress it into the smallest searchable core:\n   - brand\n   - model / series\n   - size / spec / count\n   - key variant\n4. Search the exact or lightly simplified keyword on:\n   - 京东\n   - 拼多多\n   - 唯品会\n5. Stay in browser-driven flows only. Do not call site APIs, hidden JSON endpoints, app-only interfaces, or unofficial scrapers.\n6. Extract the top relevant visible results from each site.\n7. Normalize each visible price into a comparable basis before ranking.\n8. Score same-item comparability before judging price.\n9. Apply category playbook gates before recommending a lower-price platform.\n10. Decide with recommendation strength: `强推荐`, `弱推荐`, `仅供参考`, or `无法判断`.\n11. End with a concrete recommendation: buy on which platform, stay with Taobao, wait, or avoid for now.\n\n## Input Rules\n\nRequire a product identity as input.\n\nPrefer one of these inputs:\n- Taobao product link, screenshot, title, or visible listing details\n- precise product name\n- brand + model + spec\n- product name plus intended use, if there are multiple variants\n- Taobao visible price or expected Taobao budget, if the user wants a switch / stay decision\n- category cue, such as phone, skincare, baby formula, shoes, snacks, tissue, appliance, or supplement\n- buying priority, such as lowest price, official/authenticity, fast delivery, easy returns, warranty/invoice, exact color/size, or competitor research\n\nIf the product name is too broad, ask one short follow-up to narrow it, for example:\n- brand\n- model\n- size/specification\n- package count\n- flavor/color/version\n- whether the user prioritizes lowest price, authenticity/after-sales, or delivery speed\n\nGood inputs:\n- `Apple AirPods Pro 2`\n- `维达抽纸 3层 100抽 24包`\n- `耐克 Air Zoom Pegasus 41 男款`\n- `这个淘宝价 129，帮我看看京东拼多多唯品会有没有更值的同款`\n\nWeak inputs that need clarification:\n- `纸巾`\n- `耳机`\n- `运动鞋`\n\nIf the user pastes a very long Taobao title, do not ask them to rewrite it unless it is truly ambiguous. You should clean and normalize it yourself first.\n\n## Taobao Baseline Rules\n\nTreat Taobao as the baseline only when baseline evidence exists.\n\nCapture when visible or user-provided:\n- Taobao title\n- Taobao displayed price\n- coupon-after, membership, or promo wording\n- spec / version / count / packaging\n- seller/store type\n- shipping or delivery note\n- URL or screenshot context\n\nIf the baseline is only a title:\n- compare competitor platforms against the title identity\n- say `淘宝基准价未提供，本次只比较竞品平台可见结果`\n- avoid saying `值得从淘宝换平台` unless the user later provides Taobao price or visible Taobao evidence\n\nIf the user provides a Taobao price:\n- call it `用户提供的淘宝基准价`\n- compare it to normalized competitor prices\n- flag that final payable price may change with address, coupon eligibility, account status, and stock\n\n## Browser Execution Rules\n\n- Prefer the isolated OpenClaw browser unless the user explicitly asks to use their Chrome tab.\n- Start with one tab per site when practical.\n- Re-snapshot after navigation or major DOM changes.\n- If a site shows login walls, anti-bot interstitials, region prompts, or app-download overlays, use the visible web result if possible and mention the limitation.\n- If a site blocks access completely, report it instead of trying to bypass it.\n- Do not fabricate missing prices.\n- Prefer visible search/listing pages over deep product pages when one platform is unstable.\n- Capture enough evidence to justify the recommendation, not just enough to fill a table.\n\n## Search Targets\n\nUse the standard web search pages when possible:\n\n- 京东: search for the product name on jd.com\n- 拼多多: search for the product name on pinduoduo.com or the visible web listing/search experience available in browser\n- 唯品会: search for the product name on vip.com\n\nIf direct site search is unstable in browser, use a public search engine query constrained to the site, then open the most relevant visible result. Example pattern:\n- `site:jd.com 商品名`\n- `site:pinduoduo.com 商品名`\n- `site:vip.com 商品名`\n\nStill use browser navigation for the actual evidence collection.\n\n## Category And Preference Routing\n\nRead `references/category-playbooks.md` when the product category is obvious or when price, trust, warranty, freshness, sizing, or authenticity can change the decision.\n\nClassify the category before final recommendation:\n- `数码/家电`: model, version, warranty, invoice, self-operated/official channel, installation, trade-in, refurbished/open-box risk.\n- `美妆/个护`: official channel, batch/expiry, sample/trial size, authenticity, sealed packaging, return limits.\n- `食品/日用品`: unit price, pack count, shelf life, shipping threshold, heavy-item delivery, commodity substitutability.\n- `服饰/鞋包`: exact color/size/style code, season/version, authenticity, return convenience, stock by size.\n- `母婴/健康/安全`: official/authorized channel, registration/approval when relevant, expiry, warranty, return policy, safety risk.\n- `图书/文具/低风险标品`: ISBN/model/count, shipping, bundle contents, seller reliability, unit price.\n- `不确定/混合`: say which playbook you used and why.\n\nInfer buying posture:\n- `低价优先`: maximize normalized savings, but only after match and risk gates pass.\n- `正品/售后优先`: prefer official, self-operated, flagship, authorized, invoice, warranty, and return clarity.\n- `速度优先`: prefer visible delivery promise and stable fulfillment even if not the cheapest.\n- `精确款优先`: prioritize exact variant, size, color, model, batch, or package count.\n- `研究优先`: broaden to near matches and substitutes, but clearly separate them from exact matches.\n\nUse category-specific price-delta thresholds:\n- For high-risk or warranty-heavy goods, require a meaningful saving before recommending a weaker channel.\n- For low-risk commodities, a clear normalized unit-price win can justify switching more easily.\n- If the category playbook conflicts with the lowest visible price, the playbook wins.\n\n## Matching Score\n\nTreat listings as comparable only when the core attributes align.\n\nScore each candidate out of 100:\n- Brand: 25\n- Product line / model / generation: 25\n- Variant and must-match specs: 20\n- Size / weight / count / packaging: 15\n- Seller/channel trust when relevant: 10\n- Evidence completeness: 5\n\nUse these match bands:\n- `高` / 85-100: same item or same SKU-equivalent listing; price comparison can drive the recommendation.\n- `中` / 65-84: near match with one explainable difference; include it with caveats, and make only a weak recommendation unless the tradeoff is obvious.\n- `低` / below 65: reference or substitute only; do not call it cheaper than Taobao or cheaper than another exact match.\n\nForce the match band to `低` when any critical mismatch appears:\n- different model, generation, storage, capacity, size, flavor, color, or pack count\n- refurbished / open-box / parallel import when the baseline is standard new domestic stock\n- trial/sample size compared with regular size\n- unclear warranty or authenticity for safety-sensitive or high-counterfeit-risk categories\n- missing enough information to confirm the listing is the same item\n\nApply category overrides:\n- For high-counterfeit, safety-sensitive, or warranty-heavy categories, reduce the score when official/self-operated/authorized evidence is missing.\n- For low-risk commodities, do not over-penalize seller/channel differences when the spec, count, and unit basis are clear.\n- For apparel and shoes, exact size, color, style code, and return policy can matter as much as headline model name.\n- For beauty, baby, food, and supplements, expiry, batch, sealed packaging, and authenticity cues can be critical match evidence.\n\nIf there is no close match on a platform:\n- say `未找到足够接近的同款`\n- optionally include the nearest visible alternative, clearly labeled as `近似款` or `替代款`\n\n## Price Normalization Rules\n\nDo not rank prices until they are normalized.\n\nFor each candidate, separate these fields when visible:\n- `标价`: the headline displayed price\n- `可见到手价`: coupon-after, subsidy, member, or promo price, only when the condition is visible\n- `优惠条件`: coupon, membership, group-buy, payment method, limited-time sale, regional subsidy, or account qualification\n- `单位基准`: per item, per pack, per 100g, per ml, per sheet, per pair, or another meaningful unit\n- `运费/门槛`: shipping fee, free-shipping threshold, installation fee, or delivery limitation\n\nUse the most conservative comparable price:\n- Prefer unconditional visible price when coupon eligibility is unclear.\n- Use coupon-after price only when the page makes the condition visible and likely attainable.\n- Keep member-only, group-buy, subsidy, bank-card, and region-dependent prices separate from normal prices.\n- For multi-pack goods, compare unit price and total package value, not only headline price.\n- For electronics, apparel, beauty, and branded goods, do not let a lower price override warranty, channel, authenticity, or version mismatch.\n- Use the category playbook's preferred unit basis when available.\n\nNever recommend a cheaper platform when the cheaper listing is only cheaper because of:\n- lower specification\n- different quantity or packaging\n- unclear seller trust\n- member-only or coupon-after price not available to most users\n- group-buy requirement the user may not want\n- missing shipping, installation, warranty, or invoice cost\n\n## Decision Rules\n\nUse this order of judgment:\n\n1. Is the candidate the same item or only a near match\n2. Is the visible price directly comparable after normalization\n3. Does the category playbook allow price to dominate, or should trust/warranty/authenticity/freshness/sizing dominate\n4. Is the seller/store trust level similar\n5. Are coupon, membership, subsidy, or group-buy conditions required\n6. Are shipping speed, warranty, invoice, return rights, and after-sales meaningfully different\n7. Does the user's likely preference favor lowest price, trust, delivery speed, exact variant availability, or broad research\n\nRecommendation strength:\n- `强推荐`: high match, clear normalized price, trustworthy seller/channel, and the advantage is material.\n- `弱推荐`: high or medium match, but with one meaningful caveat such as coupon dependency, seller difference, delivery uncertainty, category risk, or missing Taobao baseline price.\n- `仅供参考`: partial evidence, noisy results, medium/low match, or blocked platform pages.\n- `无法判断`: no comparable listings, missing baseline for the user question, or evidence too incomplete to support a purchase decision.\n\nDo not use one universal price threshold:\n- A 5-8% saving may be weak for a phone if warranty/channel is worse.\n- A 10-20% normalized unit-price saving may matter for low-risk daily goods.\n- Any saving can be irrelevant if the exact size, color, model, expiry, or warranty does not match.\n\nIf Taobao is not actually the best option, say so directly. If none of the results are truly comparable, explicitly say `当前不适合做强结论`.\n\n## What To Capture Per Site\n\nCapture only information visible on the page. Prefer the first 1-3 relevant results.\n\nFor each selected listing, collect when visible:\n- platform\n- title\n- displayed price\n- visible final/promo price and condition\n- unit price or normalized price basis, when applicable\n- package/specification\n- store/seller name\n- delivery or shipping note\n- URL\n- match score and band\n- category and buying-posture note when it changes the conclusion\n- confidence: high / medium / low\n- note about why it matches or why it is only approximate\n\nAlso capture, when visible and relevant:\n- official/self-operated/flagship indicator\n- coupon or subsidy dependency\n- group-buy requirement\n- delivery speed or shipping promise\n- return, warranty, invoice, or authenticity guarantee\n\nEvidence quality gate:\n- A candidate should include at least title, displayed price, spec/version, and URL to drive a recommendation.\n- If two or more of those fields are missing, downgrade confidence to low.\n- Low-confidence candidates can appear in the table but cannot be the sole basis for `强推荐`.\n\n## Output Format\n\nReturn a decision first, then the evidence table.\n\nStart with a short verdict block:\n\n- `淘宝基准`: visible/user-provided price, or `淘宝基准价未提供`\n- `品类策略`: category playbook used and why\n- `购买偏好`: inferred or user-provided priority\n- `推荐平台`: platform or `暂不建议换平台`\n- `推荐强度`: `强推荐` / `弱推荐` / `仅供参考` / `无法判断`\n- `最低可见可比价`: platform + normalized price basis, only among comparable items\n- `值不值得换平台`: yes / no / depends, with one sentence\n- `主要原因`\n- `风险点`\n- `核验时间`: include date/time or say browser-visible at time of checking\n\nThen return a concise comparison table and short notes.\n\nUse a table like this:\n\n| 平台 | 商品标题 | 标价 | 可见到手价/条件 | 单位价/基准 | 规格/版本 | 店铺 | 匹配分 | 匹配度 | 备注 |\n|---|---|---:|---|---|---|---|---:|---|---|\n| 淘宝 | ... | ¥... | ... | ... | ... | ... | 基准 | 基准 | 用户提供/页面可见 |\n| 京东 | ... | ¥... | ... | ... | ... | ... | 92 | 高 | 同品牌同规格 |\n| 拼多多 | ... | ¥... | ... | ... | ... | ... | 78 | 中 | 规格接近，包装不同 |\n| 唯品会 | ... | ¥... | ... | ... | ... | ... | 52 | 低 | 仅找到近似款 |\n\nThen add:\n- `最低可见价`: platform and raw visible price\n- `最低可比价`: platform and normalized comparable price\n- `可比性判断`: high / medium / low, with reason\n- `证据完整性`: state whether each recommendation-driving row has title, price, spec/version, and URL\n- `品类判断`: why this category should prioritize price, trust, delivery, warranty, freshness, authenticity, or exact variant\n- `风险提示`: differences in package, seller, promo timing, membership price, shipping, warranty, invoice, or coupon requirements\n- `购买建议`: 直接买 / 可等等 / 只建议在某平台买 / 暂不建议下单\n- `下单前核对`: final payable price, address-based delivery, coupon eligibility, stock, warranty, and return policy\n\n## Interpretation Rules\n\n- Do not claim a platform is cheaper unless the compared items are materially comparable.\n- Separate `标价` from coupon-after price when the page makes that distinction.\n- Mention when a price may depend on membership, flash sale, subsidy, payment method, group-buy, or region.\n- If search results are noisy, prefer accuracy over completeness.\n- If the Taobao baseline price is missing, do not say the user should switch away from Taobao; say which competitor has the best visible comparable offer.\n- If the Taobao baseline is not actually the best option, say so directly.\n- If category-specific risk outweighs the price difference, recommend staying with the safer channel or waiting.\n- If none of the results are truly comparable, explicitly say `当前不适合做强结论`.\n- Do not issue `强推荐` unless the recommendation-driving listing includes at least title, price, spec/version, URL, and enough seller/channel evidence for the category risk.\n- If the user seems purchase-ready, optimize the answer for actionability: where to buy, what to verify before paying, and what tradeoff they are accepting.\n\n## Example Requests\n\n- `帮我查一下“德芙黑巧克力 84g”在京东、拼多多、唯品会的价格`\n- `对比一下“iPhone 16 Pro 256GB”在几个平台上的可见报价`\n- `把这个淘宝商品名拿去京东、拼多多、唯品会搜同款，做个价格表`\n- `这个淘宝商品有没有更便宜但靠谱的平台`\n- `帮我判断这件商品有没有必要从淘宝换到京东买`\n- `淘宝价 129，这个商品换平台买值不值`\n- `别只比价，也告诉我哪个平台更值得下单`\n\n## Failure Handling\n\nIf one or more sites cannot be accessed or searched reliably, still return a partial result and list:\n- which site failed\n- what was attempted\n- whether the failure was due to login wall, anti-bot page, timeout, app-only flow, or missing web search results\n\nIf evidence is partial, downgrade recommendation strength.\n\nDo not fill missing evidence with guesses. Use:\n- `未见明确价格`\n- `未见规格`\n- `未见店铺信息`\n- `无法确认同款`\n\n## Resource\n\n- Read `references/site-notes.md` when you need execution reminders for JD, Pinduoduo, and Vipshop search behavior, evidence standards, normalized pricing, and recommendation gates.\n- Read `references/category-playbooks.md` when category, user preference, authenticity, warranty, freshness, sizing, safety, or platform fit can change the recommendation.\n\n\n## P1 Safety Boundaries\n\n- Do not enter credentials, SMS codes, passwords, CAPTCHA, identity checks, addresses, or payment details for the user.\n- Do not submit orders, click checkout, click final confirmation, or initiate payment.\n- Use browser-visible or user-provided information only; final price, stock, delivery, coupons, and after-sales terms must be rechecked by the user before purchase.\n\nFile v2.1.1:README.md\n\n# Taobao Competitor Analyzer\n\nCross-platform shopping decision skill for Taobao users.\n\nThis skill checks the same or closest-matching product on:\n- JD.com\n- Pinduoduo\n- Vipshop\n\nThen it answers the question users actually care about:\n\n`Should I keep buying this on Taobao, switch platforms, wait, or avoid this deal?`\n\n## What It Does\n\n- Compares visible prices across major Chinese marketplaces\n- Uses Taobao as the baseline when a Taobao price, link, screenshot, or visible listing is available\n- Matches brand, model, spec, count, and packaging before comparing price\n- Scores same-item confidence so near matches do not masquerade as cheaper exact matches\n- Normalizes headline price, coupon-after price, member price, subsidy price, group-buy price, shipping, and unit price\n- Applies category playbooks so phones, skincare, baby goods, shoes, snacks, and paper towels are judged differently\n- Flags near-match risk instead of pretending similar items are identical\n- Weighs seller trust, shipping, warranty, invoice, and after-sales guarantees\n- Returns a recommendation strength: strong, weak, reference-only, or cannot judge\n\n## v1.1.0 Highlights\n\n- Taobao baseline handling: no baseline price means no unsupported \"switch away from Taobao\" claim\n- 100-point match scoring across brand, model, specs, quantity, channel trust, and evidence completeness\n- Price normalization for unit price, coupon dependency, membership, subsidy, group-buy, and shipping caveats\n- Evidence quality gate for title, price, spec/version, and URL before a listing can drive a strong recommendation\n- Expanded verdict block with recommendation strength, lowest comparable price, risks, and final checks before purchase\n\n## v1.2.0 Highlights\n\n- Category-aware playbooks for electronics/appliances, beauty/personal care, food/daily goods, apparel/shoes/bags, baby/health/safety, and low-risk standard goods\n- Buying-posture routing for lowest price, authenticity/after-sales, delivery speed, exact variant, and research-first workflows\n- Platform-fit judgment so JD, PDD, and Vipshop are weighed differently by category instead of by headline price alone\n- Category-specific price-delta thresholds: low-risk commodities can switch on unit-price wins, while warranty/authenticity categories need stronger evidence\n- New `品类策略` and `购买偏好` fields in the verdict block\n\n## Best Use Cases\n\n- Compare a Taobao listing with JD / PDD / Vipshop\n- Check whether a \"cheaper\" result is actually the same SKU\n- Decide if it is worth switching platforms\n- Do fast competitor research for consumer products\n- Find the lowest visible comparable price with caveats\n- Separate normal price from coupon, member, subsidy, and group-buy conditions\n- Decide whether a low price is worth the category-specific risk\n\n## Example Prompts\n\n- `帮我查这个淘宝商品在京东、拼多多、唯品会有没有同款`\n- `淘宝价 129，这个商品换到京东买值不值`\n- `这款护肤品拼多多便宜很多，风险大不大`\n- `这双鞋唯品会有近似款，能不能买`\n- `别只比价，告诉我哪个平台更值得买`\n- `Compare this Taobao product with JD, PDD, and Vipshop and recommend where to buy`\n\n## Output Style\n\nThe skill is optimized to return:\n- Taobao baseline status\n- recommended platform\n- recommendation strength\n- lowest visible comparable price\n- normalized unit/condition basis\n- category strategy and inferred buying priority\n- whether the comparison is apples-to-apples\n- major risks and caveats\n- a direct buy / wait / avoid recommendation\n\n## Positioning\n\nThis is not a generic shopping browser helper.\n\nIt is a focused purchase-decision skill for users who want:\n- faster price comparison\n- fewer fake \"cheap\" matches\n- clearer marketplace tradeoff analysis\n- conservative recommendations when evidence is incomplete\n- category-aware buying advice rather than one-size-fits-all price sorting\n\nFile v2.1.1:_meta.json\n\n{\n  \"ownerId\": \"kn77zzg9p845zanvy6vrf76k7d81mcnm\",\n  \"slug\": \"taobao-competitor-analyzer\",\n  \"version\": \"2.1.1\",\n  \"publishedAt\": 1782214405857\n}\n\nFile v2.1.1:references/category-playbooks.md\n\n# Category Playbooks\n\n## Goal\n\nUse these playbooks after normalizing the Taobao baseline and before making the final recommendation. The goal is to avoid one-size-fits-all price sorting. A lower price means different things for phones, skincare, baby goods, shoes, snacks, and paper towels.\n\n## How To Use\n\n1. Classify the product into the closest category.\n2. Infer the user's buying posture: low price, trust/after-sales, delivery speed, exact variant, or research.\n3. Apply the category's must-check attributes and risk gates.\n4. Decide whether the price gap is large enough to overcome category risk.\n5. State the playbook in the verdict as `品类策略`.\n\nIf the category is unclear, write the assumption and keep recommendation strength at `弱推荐` or lower.\n\n## Buying Postures\n\n### Low Price First\n\n- Use normalized unit price after matching gates pass.\n- Prefer unconditional visible prices over conditional coupon/member/group-buy prices.\n- Strong recommendations still need high match and enough seller/channel evidence.\n\n### Trust And After-Sales First\n\n- Prefer official, self-operated, flagship, or authorized channels.\n- Treat invoice, warranty, returns, and authenticity evidence as decision drivers.\n- A cheaper weak-channel listing usually becomes `弱推荐` or `仅供参考`.\n\n### Speed First\n\n- Prefer visible delivery promise, local stock, self-operated logistics, and stable fulfillment.\n- Downgrade group-buy, pre-sale, limited stock, and unclear shipping.\n- If delivery is address-dependent and not visible, include it in `下单前核对`.\n\n### Exact Variant First\n\n- Prioritize exact model, color, size, flavor, version, pack count, batch, and bundle contents.\n- Do not treat near matches as cheaper exact matches.\n- For apparel, shoes, beauty, and electronics, variant mismatch usually blocks `强推荐`.\n\n### Research First\n\n- Include exact matches, near matches, and substitutes in separate groups.\n- Do not rank substitutes against exact matches as if they are identical.\n- Use this mode for marketplace research, category scans, and competitor mapping.\n\n## Electronics And Appliances\n\nExamples: phones, headphones, laptops, cameras, routers, appliances, smart devices.\n\nMust check:\n- exact model, generation, storage/capacity, color, region/version, bundle, and warranty\n- new vs refurbished/open-box/parallel import\n- official/self-operated/authorized channel\n- invoice, warranty, installation, return policy, trade-in, and delivery timing\n\nPlatform fit:\n- JD is often stronger when self-operated, official, invoice, warranty, fast delivery, or installation matters.\n- PDD can be considered when subsidy/official evidence is visible and the SKU is exact.\n- Vipshop is usually weaker for exact electronics coverage unless the listing is clearly official and exact.\n\nRecommendation gate:\n- Do not recommend a weaker channel for small savings.\n- Require high match and clear warranty/channel evidence for `强推荐`.\n- If savings are modest and JD/Taobao official support is stronger, recommend staying with the safer channel or waiting.\n\n## Beauty And Personal Care\n\nExamples: skincare, perfume, makeup, haircare, oral care, personal care devices.\n\nMust check:\n- exact product line, volume, shade, scent, set contents, version, and packaging\n- official/flagship/authorized seller cues\n- batch, expiry, sealed packaging, import/domestic version, sample/trial size\n- return limits and authenticity guarantee\n\nPlatform fit:\n- Tmall/Taobao flagship, JD official/self-operated, and brand official channels are stronger trust signals.\n- Vipshop can be attractive for branded discount inventory when exact variant and channel confidence are visible.\n- PDD low prices need extra caution unless official/subsidy/channel evidence is clear.\n\nRecommendation gate:\n- Do not let a low price override unclear authenticity, sample size, expiry, or seller trust.\n- For branded skincare and perfume, weak-channel savings should usually be `仅供参考`.\n- If exact shade/volume/set differs, mark as near match or substitute.\n\n## Food And Daily Goods\n\nExamples: snacks, drinks, rice/oil, tissue, detergent, pet food, household consumables.\n\nMust check:\n- unit count, weight, volume, flavor, pack count, expiration/shelf life, and packaging\n- unit price basis such as per 100g, per bottle, per pack, per sheet, or per item\n- shipping fee, free-shipping threshold, heavy-item delivery, and regional stock\n- brand and seller reliability for food, pet food, and ingestible items\n\nPlatform fit:\n- PDD can win on low-risk commodities when spec, count, and unit basis are clear.\n- JD can win when speed, heavy-item delivery, after-sales, or self-operated reliability matters.\n- Vipshop is less central unless it has exact branded inventory or bundle discounts.\n\nRecommendation gate:\n- A clear normalized unit-price win can justify switching for low-risk daily goods.\n- For food, pet food, or anything ingested, expiry and seller reliability still matter.\n- Do not compare different pack counts without unit price.\n\n## Apparel, Shoes, And Bags\n\nExamples: clothing, sneakers, sports shoes, bags, accessories.\n\nMust check:\n- brand, style code/model, size, color, gender, season/version, material, and bundle/accessories\n- official/authorized seller when counterfeiting risk is meaningful\n- return/exchange policy, size availability, stock, and authenticity guarantee\n\nPlatform fit:\n- Vipshop can be strong for branded discount inventory when size/color/style code match exactly.\n- JD/Tmall/official channels are stronger when authenticity and returns matter.\n- PDD and generic Taobao listings require caution for branded goods unless evidence is strong.\n\nRecommendation gate:\n- Size/color mismatch usually means near match, not same item.\n- If returns are unclear for apparel/shoes, downgrade recommendation strength.\n- Do not recommend a cheaper listing for branded shoes/bags without strong authenticity evidence.\n\n## Baby, Health, And Safety\n\nExamples: infant formula, baby products, supplements, medicine-adjacent goods, helmets, batteries, chargers, appliances with safety implications, medical devices.\n\nMust check:\n- official/authorized seller, registration/approval where relevant, batch, expiry, standard/certification, warranty, and return policy\n- exact model/spec and safety certification\n- do not provide medical claims or health guarantees\n\nPlatform fit:\n- Prefer official, self-operated, flagship, authorized, or brand channels.\n- Treat very low prices from unclear sellers as risk signals.\n- PDD or marketplace third-party listings need unusually strong evidence to be more than `仅供参考`.\n\nRecommendation gate:\n- Safety and authenticity override price.\n- Do not issue `强推荐` for unclear-channel baby/health/safety goods.\n- If evidence is weak, recommend official/self-operated channels or waiting.\n\n## Books, Stationery, And Low-Risk Standard Goods\n\nExamples: books, notebooks, pens, cables with low risk, small office supplies, simple accessories.\n\nMust check:\n- ISBN/model, edition, count, color, size, bundle contents, shipping, and seller reliability\n- for books, distinguish正版,影印,二手,预售,套装, and different editions\n\nPlatform fit:\n- PDD and Taobao can be reasonable for low-risk standard goods when listing details are clear.\n- JD can be better for fast shipping, invoices, and standardized fulfillment.\n- Vipshop is usually relevant only for branded stationery or limited discount inventory.\n\nRecommendation gate:\n- Price can matter more once edition/model/count are confirmed.\n- Do not compare different editions, bundle counts, or used/new condition as exact matches.\n\n## Unknown Or Mixed Category\n\nUse this when the item does not fit cleanly.\n\nRules:\n- State the assumed category.\n- Use the strictest relevant risk gate among possible categories.\n- Keep recommendation at `弱推荐` or lower unless evidence is very strong.\n- Ask one short follow-up if the category would materially change the purchase advice.\n\nFile v2.1.1:references/site-notes.md\n\n# Site Notes\n\n## Goal\n\nUse browser-visible marketplace pages to collect price evidence for the same or closest-matching product on JD, Pinduoduo, and Vipshop, then turn that evidence into a purchase decision against the Taobao baseline when available.\n\n## Common Reminders\n\n- Prefer visible desktop web pages.\n- Use exact product names first, then a lightly simplified query if results are sparse.\n- Compare like-for-like products only.\n- Keep the evidence chain simple: search page -> listing -> visible title/price/spec.\n- Do not use APIs, hidden JSON endpoints, or scraping shortcuts outside the browser tool.\n- Price is not enough; capture trust, conditions, and caveats.\n- The final answer should help the user choose, not just browse.\n- Read `category-playbooks.md` when the product category changes how much price should matter.\n\n## Baseline Discipline\n\n- Treat Taobao as the baseline only when a Taobao price, URL, screenshot, or user-provided listing detail exists.\n- If the user provides only a Taobao title, search against that identity but write `淘宝基准价未提供`.\n- Do not claim the user should switch away from Taobao unless Taobao's visible or user-provided price is known.\n- When the user provides the Taobao price, label it as user-provided unless verified in browser.\n\n## Matching Score Reminders\n\nUse the 100-point score from `SKILL.md`:\n- brand: 25\n- model / line / generation: 25\n- variant and must-match specs: 20\n- size / weight / count / packaging: 15\n- seller or channel trust: 10\n- evidence completeness: 5\n\nInterpretation:\n- 85-100: high, can drive recommendation\n- 65-84: medium, caveated recommendation only\n- below 65: low, reference/near-match only\n\nDowngrade to low when the result has a critical mismatch such as different generation, capacity, count, refurbished status, unclear warranty, trial size, or unverified authenticity in a risk-sensitive category.\n\n## Price Normalization Reminders\n\nBefore ranking, separate:\n- headline/list price\n- visible final or coupon-after price\n- coupon/member/subsidy/group-buy/payment/region conditions\n- unit basis such as per pack, per sheet, per 100g, per ml, per pair, or per item\n- shipping, free-shipping threshold, installation, warranty, invoice, or delivery constraints\n\nUse conservative comparison:\n- Prefer unconditional visible price when promo eligibility is unclear.\n- Treat member-only, group-buy, payment, and regional subsidy prices as conditional.\n- For multi-pack daily goods, compare unit price and total quantity.\n- For electronics, beauty, apparel, infant goods, medical/safety goods, and branded products, keep authenticity, warranty, and channel trust ahead of headline price.\n\n## Category Shortcut\n\nUse `references/category-playbooks.md` to decide which risk dominates:\n- warranty and invoice for electronics/appliances\n- authenticity, batch, and expiry for beauty/personal care\n- unit price, pack count, and freshness for food/daily goods\n- size, color, style code, and return policy for apparel/shoes/bags\n- official authorization and safety evidence for baby/health/safety products\n- ISBN/model/count and shipping for books/stationery/low-risk standard goods\n\nWhen category risk and headline price disagree, mention the conflict and let category risk control recommendation strength.\n\n## JD\n\n- Usually supports standard web search and listing pages well.\n- Prefer self-operated or official flagship listings when multiple near-identical results exist.\n- Watch for coupon text, plus/member pricing, trade-in, bank/payment offers, and promotional banners.\n- Distinguish list price from final promo price if both are shown.\n- Treat 京东自营 and official flagship stores as stronger trust signals when prices are close.\n- For electronics, appliances, beauty, infant goods, and regulated products, do not recommend a weaker channel solely because it is cheaper.\n\n## Pinduoduo\n\n- Web results may be noisier than JD.\n- Watch for subsidy labels, group-buy wording, coupon claims, and strong promo framing.\n- Matching confidence should be reduced when the seller, spec, or packaging is unclear.\n- If the browser experience is limited, use a public search engine with site restriction, then open the visible result page.\n- If the low price depends on 拼团 or 百亿补贴, call that out explicitly in the final recommendation.\n- Reduce recommendation strength when seller trust is unclear, even if the price looks excellent.\n- A very low PDD price should be treated as a lead to verify, not as a conclusion, unless match and seller evidence are strong.\n\n## Vipshop\n\n- Results may lean toward branded discount inventory and variant-specific listings.\n- Pay attention to size/color/version because discount channels often surface adjacent variants.\n- If the exact match is missing, mark the result as `近似款` instead of treating it as the same SKU.\n- Vipshop often wins on branded discounts but loses on exact spec coverage; do not overstate comparability.\n- Treat limited stock, size gaps, color mismatch, and flash-sale timing as decision caveats.\n\n## Evidence Standard\n\nFor each platform, aim to capture:\n- title\n- displayed price\n- visible final price or promo condition\n- variant/specification\n- seller/store if visible\n- URL\n- matching score and reason\n- a short matching note\n\nWhen visible, also capture:\n- official / self-operated / flagship status\n- coupon or membership dependency\n- subsidy / group-buy dependency\n- shipping promise\n- return / warranty / invoice / authenticity guarantee\n\nEvidence quality gate:\n- A listing should have title, price, spec/version, and URL before it can support a strong recommendation.\n- If two or more of those are missing, downgrade confidence to low.\n- Low-confidence candidates can be included, but they should not determine the final answer alone.\n\n## Output Discipline\n\nUse short, decision-friendly language.\n\nRecommended summary fields:\n- Taobao baseline status\n- category strategy and buying posture\n- lowest visible raw price\n- lowest visible comparable price\n- whether the comparison is apples-to-apples\n- major caveats: variant mismatch, coupon dependency, seller difference, shipping difference, warranty/invoice difference, regional limitation\n- recommended platform\n- recommendation strength: strong / weak / reference only / cannot judge\n- one-sentence reason the user should or should not switch away from Taobao\n- final checks before buying: final payable price, address-based delivery, coupon eligibility, stock, warranty, and return policy\n\nFile v2.1.1:skill-card.md\n\n## Description:\n\nTaobao price and competitor comparison assistant. Input a Taobao title/link or product need; compare JD, PDD, and Vipshop using browser-visible evidence, normalize prices, score same-item confidence, and output where to buy. Safe boundary: no login, no order submission, no payment.\n\nThis skill is ready for commercial/non-commercial use.\n\n## Publisher:\n\n[harrylabsj](https://clawhub.ai/user/harrylabsj)\n\n### License/Terms of Use:\n\nMIT-0\n\n## Use Case:\n\nExternal shoppers and shopping assistants use this skill to compare a Taobao product against JD, Pinduoduo, and Vipshop visible listings, normalize prices and promo conditions, and decide whether to buy, wait, avoid, or stay with Taobao.\n\n### Deployment Geography for Use:\n\nGlobal\n\n## Known Risks and Mitigations:\n\nRisk: Visible marketplace prices, coupons, stock, shipping, warranty, and return terms can change or depend on account or address.\n\nMitigation: Use the skill's output as a comparison aid and recheck final payable price, coupon eligibility, stock, delivery, warranty, and returns before buying.\n\nRisk: A lower listing price may reflect a different specification, package count, seller channel, promo condition, or near-match rather than the same item.\n\nMitigation: Require visible title, price, spec or version, and URL, normalize by unit and promo conditions, and downgrade recommendations when match or channel evidence is incomplete.\n\nRisk: Marketplace pages may show login walls, CAPTCHA, app-only flows, or other access limits.\n\nMitigation: Do not log in, solve CAPTCHA, enter personal details, submit orders, or pay; report access limits and use only visible public evidence.\n\n## Reference(s):\n\n- [Category Playbooks](references/category-playbooks.md)\n- [Site Notes](references/site-notes.md)\n- [Taobao Price Compare on ClawHub](https://clawhub.ai/harrylabsj/skills/taobao-competitor-analyzer)\n\n## Skill Output:\n\n**Output Type(s):** [Text, Markdown, Guidance]\n\n**Output Format:** [Markdown verdict block, comparison table, and concise risk notes]\n\n**Output Parameters:** [1D]\n\n**Other Properties Related to Output:** [Uses browser-visible or user-provided marketplace evidence and may include Chinese field labels.]\n\n## Skill Version(s):\n\n2.1.1 (source: server release metadata)\n\n## Ethical Considerations:\n\nUsers should evaluate whether this skill is appropriate for their environment, review any generated or modified files before relying on them, and apply their organization's safety, security, and compliance requirements before deployment.\n\nArchive v2.1.0: 5 files, 16529 bytes\n\nFiles: references/category-playbooks.md (7975b), references/site-notes.md (6507b), skill-card.md (2298b), SKILL.md (20216b), _meta.json (145b)\n\nFile v2.1.0:SKILL.md\n\n---\nname: taobao-competitor-analyzer\ndescription: \"Taobao price and competitor comparison assistant. Input a Taobao title/link or product need; compare JD, PDD, and Vipshop using browser-visible evidence, normalize prices, score same-item confidence, and output where to buy. Safe boundary: no login, no order submission, no payment.\"\n---\n# Taobao Competitor Analyzer\n\nCompare a Taobao product with the same or closest-matching listings on 京东、拼多多、唯品会 using the browser tool only. Work from a product name, Taobao title, link, screenshot, or user-provided baseline details. Return a compact purchase decision with visible price evidence, normalized comparability, match score, category-aware platform fit, recommendation strength, and risk notes.\n\nWhat makes this skill useful:\n- It compares comparable items instead of chasing misleading low prices.\n- It separates Taobao baseline price from competitor prices.\n- It normalizes price by spec, quantity, shipping, and promo conditions before ranking.\n- It scores same-item confidence before making a buying recommendation.\n- It applies category-specific risk thresholds instead of treating shoes, phones, skincare, and paper towels the same way.\n- It explains whether a lower price depends on coupons, membership, subsidy, or group-buy.\n- It ends with a clear buy / wait / avoid recommendation instead of just a table.\n\n## When To Use\n\nUse this skill when the user is effectively asking:\n- 这件淘宝商品别的平台多少钱\n- 有没有同款或更划算的平台\n- 京东 / 拼多多 / 唯品会 哪个更值得买\n- 这几个平台价格差这么多正常吗\n- 帮我做一个同款比价和购买建议\n- 这个淘宝商品换平台买值不值\n- 同样价差下哪个平台风险更低\n\nThe skill should optimize for purchase decisions, not raw data collection.\n\n## Commerce Matrix\n\nThis skill is the cross-platform comparison node in the shopping matrix.\n\nPrefer nearby skills when the task is narrower:\n- `taobao-shopping` for Taobao-only listing and seller evaluation\n- `jd-shopping` for trust-first self-operated buying\n- `pdd-shopping` for low-price and subsidy-first buying\n- `tianmao` for flagship-store and authenticity-first buying\n- `vip` for branded discount and flash-sale buying\n- `alibaba-shopping` when the user first needs to choose between Taobao, Tmall, and 1688\n\n## Workflow\n\n1. Normalize the Taobao baseline.\n   - Extract product identity: brand, model / series, variant, size / spec / count, color / flavor / version, packaging, and must-match attributes.\n   - Record the Taobao price only if the user provides it or it is browser-visible.\n   - If no Taobao price is available, write `淘宝基准价未提供` and do not imply whether switching platforms saves money versus Taobao.\n2. Classify the product category and buying posture.\n   - Pick the closest category from `references/category-playbooks.md`.\n   - Infer whether the user is price-first, trust-first, speed-first, exact-variant-first, or research-first.\n   - If the category or posture would change the recommendation and cannot be inferred, ask one short follow-up.\n3. If the input is only a Taobao-style long title, compress it into the smallest searchable core:\n   - brand\n   - model / series\n   - size / spec / count\n   - key variant\n4. Search the exact or lightly simplified keyword on:\n   - 京东\n   - 拼多多\n   - 唯品会\n5. Stay in browser-driven flows only. Do not call site APIs, hidden JSON endpoints, app-only interfaces, or unofficial scrapers.\n6. Extract the top relevant visible results from each site.\n7. Normalize each visible price into a comparable basis before ranking.\n8. Score same-item comparability before judging price.\n9. Apply category playbook gates before recommending a lower-price platform.\n10. Decide with recommendation strength: `强推荐`, `弱推荐`, `仅供参考`, or `无法判断`.\n11. End with a concrete recommendation: buy on which platform, stay with Taobao, wait, or avoid for now.\n\n## Input Rules\n\nRequire a product identity as input.\n\nPrefer one of these inputs:\n- Taobao product link, screenshot, title, or visible listing details\n- precise product name\n- brand + model + spec\n- product name plus intended use, if there are multiple variants\n- Taobao visible price or expected Taobao budget, if the user wants a switch / stay decision\n- category cue, such as phone, skincare, baby formula, shoes, snacks, tissue, appliance, or supplement\n- buying priority, such as lowest price, official/authenticity, fast delivery, easy returns, warranty/invoice, exact color/size, or competitor research\n\nIf the product name is too broad, ask one short follow-up to narrow it, for example:\n- brand\n- model\n- size/specification\n- package count\n- flavor/color/version\n- whether the user prioritizes lowest price, authenticity/after-sales, or delivery speed\n\nGood inputs:\n- `Apple AirPods Pro 2`\n- `维达抽纸 3层 100抽 24包`\n- `耐克 Air Zoom Pegasus 41 男款`\n- `这个淘宝价 129，帮我看看京东拼多多唯品会有没有更值的同款`\n\nWeak inputs that need clarification:\n- `纸巾`\n- `耳机`\n- `运动鞋`\n\nIf the user pastes a very long Taobao title, do not ask them to rewrite it unless it is truly ambiguous. You should clean and normalize it yourself first.\n\n## Taobao Baseline Rules\n\nTreat Taobao as the baseline only when baseline evidence exists.\n\nCapture when visible or user-provided:\n- Taobao title\n- Taobao displayed price\n- coupon-after, membership, or promo wording\n- spec / version / count / packaging\n- seller/store type\n- shipping or delivery note\n- URL or screenshot context\n\nIf the baseline is only a title:\n- compare competitor platforms against the title identity\n- say `淘宝基准价未提供，本次只比较竞品平台可见结果`\n- avoid saying `值得从淘宝换平台` unless the user later provides Taobao price or visible Taobao evidence\n\nIf the user provides a Taobao price:\n- call it `用户提供的淘宝基准价`\n- compare it to normalized competitor prices\n- flag that final payable price may change with address, coupon eligibility, account status, and stock\n\n## Browser Execution Rules\n\n- Prefer the isolated OpenClaw browser unless the user explicitly asks to use their Chrome tab.\n- Start with one tab per site when practical.\n- Re-snapshot after navigation or major DOM changes.\n- If a site shows login walls, anti-bot interstitials, region prompts, or app-download overlays, use the visible web result if possible and mention the limitation.\n- If a site blocks access completely, report it instead of trying to bypass it.\n- Do not fabricate missing prices.\n- Prefer visible search/listing pages over deep product pages when one platform is unstable.\n- Capture enough evidence to justify the recommendation, not just enough to fill a table.\n\n## Search Targets\n\nUse the standard web search pages when possible:\n\n- 京东: search for the product name on jd.com\n- 拼多多: search for the product name on pinduoduo.com or the visible web listing/search experience available in browser\n- 唯品会: search for the product name on vip.com\n\nIf direct site search is unstable in browser, use a public search engine query constrained to the site, then open the most relevant visible result. Example pattern:\n- `site:jd.com 商品名`\n- `site:pinduoduo.com 商品名`\n- `site:vip.com 商品名`\n\nStill use browser navigation for the actual evidence collection.\n\n## Category And Preference Routing\n\nRead `references/category-playbooks.md` when the product category is obvious or when price, trust, warranty, freshness, sizing, or authenticity can change the decision.\n\nClassify the category before final recommendation:\n- `数码/家电`: model, version, warranty, invoice, self-operated/official channel, installation, trade-in, refurbished/open-box risk.\n- `美妆/个护`: official channel, batch/expiry, sample/trial size, authenticity, sealed packaging, return limits.\n- `食品/日用品`: unit price, pack count, shelf life, shipping threshold, heavy-item delivery, commodity substitutability.\n- `服饰/鞋包`: exact color/size/style code, season/version, authenticity, return convenience, stock by size.\n- `母婴/健康/安全`: official/authorized channel, registration/approval when relevant, expiry, warranty, return policy, safety risk.\n- `图书/文具/低风险标品`: ISBN/model/count, shipping, bundle contents, seller reliability, unit price.\n- `不确定/混合`: say which playbook you used and why.\n\nInfer buying posture:\n- `低价优先`: maximize normalized savings, but only after match and risk gates pass.\n- `正品/售后优先`: prefer official, self-operated, flagship, authorized, invoice, warranty, and return clarity.\n- `速度优先`: prefer visible delivery promise and stable fulfillment even if not the cheapest.\n- `精确款优先`: prioritize exact variant, size, color, model, batch, or package count.\n- `研究优先`: broaden to near matches and substitutes, but clearly separate them from exact matches.\n\nUse category-specific price-delta thresholds:\n- For high-risk or warranty-heavy goods, require a meaningful saving before recommending a weaker channel.\n- For low-risk commodities, a clear normalized unit-price win can justify switching more easily.\n- If the category playbook conflicts with the lowest visible price, the playbook wins.\n\n## Matching Score\n\nTreat listings as comparable only when the core attributes align.\n\nScore each candidate out of 100:\n- Brand: 25\n- Product line / model / generation: 25\n- Variant and must-match specs: 20\n- Size / weight / count / packaging: 15\n- Seller/channel trust when relevant: 10\n- Evidence completeness: 5\n\nUse these match bands:\n- `高` / 85-100: same item or same SKU-equivalent listing; price comparison can drive the recommendation.\n- `中` / 65-84: near match with one explainable difference; include it with caveats, and make only a weak recommendation unless the tradeoff is obvious.\n- `低` / below 65: reference or substitute only; do not call it cheaper than Taobao or cheaper than another exact match.\n\nForce the match band to `低` when any critical mismatch appears:\n- different model, generation, storage, capacity, size, flavor, color, or pack count\n- refurbished / open-box / parallel import when the baseline is standard new domestic stock\n- trial/sample size compared with regular size\n- unclear warranty or authenticity for safety-sensitive or high-counterfeit-risk categories\n- missing enough information to confirm the listing is the same item\n\nApply category overrides:\n- For high-counterfeit, safety-sensitive, or warranty-heavy categories, reduce the score when official/self-operated/authorized evidence is missing.\n- For low-risk commodities, do not over-penalize seller/channel differences when the spec, count, and unit basis are clear.\n- For apparel and shoes, exact size, color, style code, and return policy can matter as much as headline model name.\n- For beauty, baby, food, and supplements, expiry, batch, sealed packaging, and authenticity cues can be critical match evidence.\n\nIf there is no close match on a platform:\n- say `未找到足够接近的同款`\n- optionally include the nearest visible alternative, clearly labeled as `近似款` or `替代款`\n\n## Price Normalization Rules\n\nDo not rank prices until they are normalized.\n\nFor each candidate, separate these fields when visible:\n- `标价`: the headline displayed price\n- `可见到手价`: coupon-after, subsidy, member, or promo price, only when the condition is visible\n- `优惠条件`: coupon, membership, group-buy, payment method, limited-time sale, regional subsidy, or account qualification\n- `单位基准`: per item, per pack, per 100g, per ml, per sheet, per pair, or another meaningful unit\n- `运费/门槛`: shipping fee, free-shipping threshold, installation fee, or delivery limitation\n\nUse the most conservative comparable price:\n- Prefer unconditional visible price when coupon eligibility is unclear.\n- Use coupon-after price only when the page makes the condition visible and likely attainable.\n- Keep member-only, group-buy, subsidy, bank-card, and region-dependent prices separate from normal prices.\n- For multi-pack goods, compare unit price and total package value, not only headline price.\n- For electronics, apparel, beauty, and branded goods, do not let a lower price override warranty, channel, authenticity, or version mismatch.\n- Use the category playbook's preferred unit basis when available.\n\nNever recommend a cheaper platform when the cheaper listing is only cheaper because of:\n- lower specification\n- different quantity or packaging\n- unclear seller trust\n- member-only or coupon-after price not available to most users\n- group-buy requirement the user may not want\n- missing shipping, installation, warranty, or invoice cost\n\n## Decision Rules\n\nUse this order of judgment:\n\n1. Is the candidate the same item or only a near match\n2. Is the visible price directly comparable after normalization\n3. Does the category playbook allow price to dominate, or should trust/warranty/authenticity/freshness/sizing dominate\n4. Is the seller/store trust level similar\n5. Are coupon, membership, subsidy, or group-buy conditions required\n6. Are shipping speed, warranty, invoice, return rights, and after-sales meaningfully different\n7. Does the user's likely preference favor lowest price, trust, delivery speed, exact variant availability, or broad research\n\nRecommendation strength:\n- `强推荐`: high match, clear normalized price, trustworthy seller/channel, and the advantage is material.\n- `弱推荐`: high or medium match, but with one meaningful caveat such as coupon dependency, seller difference, delivery uncertainty, category risk, or missing Taobao baseline price.\n- `仅供参考`: partial evidence, noisy results, medium/low match, or blocked platform pages.\n- `无法判断`: no comparable listings, missing baseline for the user question, or evidence too incomplete to support a purchase decision.\n\nDo not use one universal price threshold:\n- A 5-8% saving may be weak for a phone if warranty/channel is worse.\n- A 10-20% normalized unit-price saving may matter for low-risk daily goods.\n- Any saving can be irrelevant if the exact size, color, model, expiry, or warranty does not match.\n\nIf Taobao is not actually the best option, say so directly. If none of the results are truly comparable, explicitly say `当前不适合做强结论`.\n\n## What To Capture Per Site\n\nCapture only information visible on the page. Prefer the first 1-3 relevant results.\n\nFor each selected listing, collect when visible:\n- platform\n- title\n- displayed price\n- visible final/promo price and condition\n- unit price or normalized price basis, when applicable\n- package/specification\n- store/seller name\n- delivery or shipping note\n- URL\n- match score and band\n- category and buying-posture note when it changes the conclusion\n- confidence: high / medium / low\n- note about why it matches or why it is only approximate\n\nAlso capture, when visible and relevant:\n- official/self-operated/flagship indicator\n- coupon or subsidy dependency\n- group-buy requirement\n- delivery speed or shipping promise\n- return, warranty, invoice, or authenticity guarantee\n\nEvidence quality gate:\n- A candidate should include at least title, displayed price, spec/version, and URL to drive a recommendation.\n- If two or more of those fields are missing, downgrade confidence to low.\n- Low-confidence candidates can appear in the table but cannot be the sole basis for `强推荐`.\n\n## Output Format\n\nReturn a decision first, then the evidence table.\n\nStart with a short verdict block:\n\n- `淘宝基准`: visible/user-provided price, or `淘宝基准价未提供`\n- `品类策略`: category playbook used and why\n- `购买偏好`: inferred or user-provided priority\n- `推荐平台`: platform or `暂不建议换平台`\n- `推荐强度`: `强推荐` / `弱推荐` / `仅供参考` / `无法判断`\n- `最低可见可比价`: platform + normalized price basis, only among comparable items\n- `值不值得换平台`: yes / no / depends, with one sentence\n- `主要原因`\n- `风险点`\n- `核验时间`: include date/time or say browser-visible at time of checking\n\nThen return a concise comparison table and short notes.\n\nUse a table like this:\n\n| 平台 | 商品标题 | 标价 | 可见到手价/条件 | 单位价/基准 | 规格/版本 | 店铺 | 匹配分 | 匹配度 | 备注 |\n|---|---|---:|---|---|---|---|---:|---|---|\n| 淘宝 | ... | ¥... | ... | ... | ... | ... | 基准 | 基准 | 用户提供/页面可见 |\n| 京东 | ... | ¥... | ... | ... | ... | ... | 92 | 高 | 同品牌同规格 |\n| 拼多多 | ... | ¥... | ... | ... | ... | ... | 78 | 中 | 规格接近，包装不同 |\n| 唯品会 | ... | ¥... | ... | ... | ... | ... | 52 | 低 | 仅找到近似款 |\n\nThen add:\n- `最低可见价`: platform and raw visible price\n- `最低可比价`: platform and normalized comparable price\n- `可比性判断`: high / medium / low, with reason\n- `品类判断`: why this category should prioritize price, trust, delivery, warranty, freshness, authenticity, or exact variant\n- `风险提示`: differences in package, seller, promo timing, membership price, shipping, warranty, invoice, or coupon requirements\n- `购买建议`: 直接买 / 可等等 / 只建议在某平台买 / 暂不建议下单\n- `下单前核对`: final payable price, address-based delivery, coupon eligibility, stock, warranty, and return policy\n\n## Interpretation Rules\n\n- Do not claim a platform is cheaper unless the compared items are materially comparable.\n- Separate `标价` from coupon-after price when the page makes that distinction.\n- Mention when a price may depend on membership, flash sale, subsidy, payment method, group-buy, or region.\n- If search results are noisy, prefer accuracy over completeness.\n- If the Taobao baseline price is missing, do not say the user should switch away from Taobao; say which competitor has the best visible comparable offer.\n- If the Taobao baseline is not actually the best option, say so directly.\n- If category-specific risk outweighs the price difference, recommend staying with the safer channel or waiting.\n- If none of the results are truly comparable, explicitly say `当前不适合做强结论`.\n- If the user seems purchase-ready, optimize the answer for actionability: where to buy, what to verify before paying, and what tradeoff they are accepting.\n\n## Example Requests\n\n- `帮我查一下“德芙黑巧克力 84g”在京东、拼多多、唯品会的价格`\n- `对比一下“iPhone 16 Pro 256GB”在几个平台上的可见报价`\n- `把这个淘宝商品名拿去京东、拼多多、唯品会搜同款，做个价格表`\n- `这个淘宝商品有没有更便宜但靠谱的平台`\n- `帮我判断这件商品有没有必要从淘宝换到京东买`\n- `淘宝价 129，这个商品换平台买值不值`\n- `别只比价，也告诉我哪个平台更值得下单`\n\n## Failure Handling\n\nIf one or more sites cannot be accessed or searched reliably, still return a partial result and list:\n- which site failed\n- what was attempted\n- whether the failure was due to login wall, anti-bot page, timeout, app-only flow, or missing web search results\n\nIf evidence is partial, downgrade recommendation strength.\n\nDo not fill missing evidence with guesses. Use:\n- `未见明确价格`\n- `未见规格`\n- `未见店铺信息`\n- `无法确认同款`\n\n## Resource\n\n- Read `references/site-notes.md` when you need execution reminders for JD, Pinduoduo, and Vipshop search behavior, evidence standards, normalized pricing, and recommendation gates.\n- Read `references/category-playbooks.md` when category, user preference, authenticity, warranty, freshness, sizing, safety, or platform fit can change the recommendation.\n\n\n## P1 Safety Boundaries\n\n- Do not enter credentials, SMS codes, passwords, CAPTCHA, identity checks, addresses, or payment details for the user.\n- Do not submit orders, click checkout, click final confirmation, or initiate payment.\n- Use browser-visible or user-provided information only; final price, stock, delivery, coupons, and after-sales terms must be rechecked by the user before purchase.\n\nFile v2.1.0:_meta.json\n\n{\n  \"ownerId\": \"kn77zzg9p845zanvy6vrf76k7d81mcnm\",\n  \"slug\": \"taobao-competitor-analyzer\",\n  \"version\": \"2.1.0\",\n  \"publishedAt\": 1781663787793\n}\n\nFile v2.1.0:references/category-playbooks.md\n\n# Category Playbooks\n\n## Goal\n\nUse these playbooks after normalizing the Taobao baseline and before making the final recommendation. The goal is to avoid one-size-fits-all price sorting. A lower price means different things for phones, skincare, baby goods, shoes, snacks, and paper towels.\n\n## How To Use\n\n1. Classify the product into the closest category.\n2. Infer the user's buying posture: low price, trust/after-sales, delivery speed, exact variant, or research.\n3. Apply the category's must-check attributes and risk gates.\n4. Decide whether the price gap is large enough to overcome category risk.\n5. State the playbook in the verdict as `品类策略`.\n\nIf the category is unclear, write the assumption and keep recommendation strength at `弱推荐` or lower.\n\n## Buying Postures\n\n### Low Price First\n\n- Use normalized unit price after matching gates pass.\n- Prefer unconditional visible prices over conditional coupon/member/group-buy prices.\n- Strong recommendations still need high match and enough seller/channel evidence.\n\n### Trust And After-Sales First\n\n- Prefer official, self-operated, flagship, or authorized channels.\n- Treat invoice, warranty, returns, and authenticity evidence as decision drivers.\n- A cheaper weak-channel listing usually becomes `弱推荐` or `仅供参考`.\n\n### Speed First\n\n- Prefer visible delivery promise, local stock, self-operated logistics, and stable fulfillment.\n- Downgrade group-buy, pre-sale, limited stock, and unclear shipping.\n- If delivery is address-dependent and not visible, include it in `下单前核对`.\n\n### Exact Variant First\n\n- Prioritize exact model, color, size, flavor, version, pack count, batch, and bundle contents.\n- Do not treat near matches as cheaper exact matches.\n- For apparel, shoes, beauty, and electronics, variant mismatch usually blocks `强推荐`.\n\n### Research First\n\n- Include exact matches, near matches, and substitutes in separate groups.\n- Do not rank substitutes against exact matches as if they are identical.\n- Use this mode for marketplace research, category scans, and competitor mapping.\n\n## Electronics And Appliances\n\nExamples: phones, headphones, laptops, cameras, routers, appliances, smart devices.\n\nMust check:\n- exact model, generation, storage/capacity, color, region/version, bundle, and warranty\n- new vs refurbished/open-box/parallel import\n- official/self-operated/authorized channel\n- invoice, warranty, installation, return policy, trade-in, and delivery timing\n\nPlatform fit:\n- JD is often stronger when self-operated, official, invoice, warranty, fast delivery, or installation matters.\n- PDD can be considered when subsidy/official evidence is visible and the SKU is exact.\n- Vipshop is usually weaker for exact electronics coverage unless the listing is clearly official and exact.\n\nRecommendation gate:\n- Do not recommend a weaker channel for small savings.\n- Require high match and clear warranty/channel evidence for `强推荐`.\n- If savings are modest and JD/Taobao official support is stronger, recommend staying with the safer channel or waiting.\n\n## Beauty And Personal Care\n\nExamples: skincare, perfume, makeup, haircare, oral care, personal care devices.\n\nMust check:\n- exact product line, volume, shade, scent, set contents, version, and packaging\n- official/flagship/authorized seller cues\n- batch, expiry, sealed packaging, import/domestic version, sample/trial size\n- return limits and authenticity guarantee\n\nPlatform fit:\n- Tmall/Taobao flagship, JD official/self-operated, and brand official channels are stronger trust signals.\n- Vipshop can be attractive for branded discount inventory when exact variant and channel confidence are visible.\n- PDD low prices need extra caution unless official/subsidy/channel evidence is clear.\n\nRecommendation gate:\n- Do not let a low price override unclear authenticity, sample size, expiry, or seller trust.\n- For branded skincare and perfume, weak-channel savings should usually be `仅供参考`.\n- If exact shade/volume/set differs, mark as near match or substitute.\n\n## Food And Daily Goods\n\nExamples: snacks, drinks, rice/oil, tissue, detergent, pet food, household consumables.\n\nMust check:\n- unit count, weight, volume, flavor, pack count, expiration/shelf life, and packaging\n- unit price basis such as per 100g, per bottle, per pack, per sheet, or per item\n- shipping fee, free-shipping threshold, heavy-item delivery, and regional stock\n- brand and seller reliability for food, pet food, and ingestible items\n\nPlatform fit:\n- PDD can win on low-risk commodities when spec, count, and unit basis are clear.\n- JD can win when speed, heavy-item delivery, after-sales, or self-operated reliability matters.\n- Vipshop is less central unless it has exact branded inventory or bundle discounts.\n\nRecommendation gate:\n- A clear normalized unit-price win can justify switching for low-risk daily goods.\n- For food, pet food, or anything ingested, expiry and seller reliability still matter.\n- Do not compare different pack counts without unit price.\n\n## Apparel, Shoes, And Bags\n\nExamples: clothing, sneakers, sports shoes, bags, accessories.\n\nMust check:\n- brand, style code/model, size, color, gender, season/version, material, and bundle/accessories\n- official/authorized seller when counterfeiting risk is meaningful\n- return/exchange policy, size availability, stock, and authenticity guarantee\n\nPlatform fit:\n- Vipshop can be strong for branded discount inventory when size/color/style code match exactly.\n- JD/Tmall/official channels are stronger when authenticity and returns matter.\n- PDD and generic Taobao listings require caution for branded goods unless evidence is strong.\n\nRecommendation gate:\n- Size/color mismatch usually means near match, not same item.\n- If returns are unclear for apparel/shoes, downgrade recommendation strength.\n- Do not recommend a cheaper listing for branded shoes/bags without strong authenticity evidence.\n\n## Baby, Health, And Safety\n\nExamples: infant formula, baby products, supplements, medicine-adjacent goods, helmets, batteries, chargers, appliances with safety implications, medical devices.\n\nMust check:\n- official/authorized seller, registration/approval where relevant, batch, expiry, standard/certification, warranty, and return policy\n- exact model/spec and safety certification\n- do not provide medical claims or health guarantees\n\nPlatform fit:\n- Prefer official, self-operated, flagship, authorized, or brand channels.\n- Treat very low prices from unclear sellers as risk signals.\n- PDD or marketplace third-party listings need unusually strong evidence to be more than `仅供参考`.\n\nRecommendation gate:\n- Safety and authenticity override price.\n- Do not issue `强推荐` for unclear-channel baby/health/safety goods.\n- If evidence is weak, recommend official/self-operated channels or waiting.\n\n## Books, Stationery, And Low-Risk Standard Goods\n\nExamples: books, notebooks, pens, cables with low risk, small office supplies, simple accessories.\n\nMust check:\n- ISBN/model, edition, count, color, size, bundle contents, shipping, and seller reliability\n- for books, distinguish正版,影印,二手,预售,套装, and different editions\n\nPlatform fit:\n- PDD and Taobao can be reasonable for low-risk standard goods when listing details are clear.\n- JD can be better for fast shipping, invoices, and standardized fulfillment.\n- Vipshop is usually relevant only for branded stationery or limited discount inventory.\n\nRecommendation gate:\n- Price can matter more once edition/model/count are confirmed.\n- Do not compare different editions, bundle counts, or used/new condition as exact matches.\n\n## Unknown Or Mixed Category\n\nUse this when the item does not fit cleanly.\n\nRules:\n- State the assumed category.\n- Use the strictest relevant risk gate among possible categories.\n- Keep recommendation at `弱推荐` or lower unless evidence is very strong.\n- Ask one short follow-up if the category would materially change the purchase advice.\n\nFile v2.1.0:references/site-notes.md\n\n# Site Notes\n\n## Goal\n\nUse browser-visible marketplace pages to collect price evidence for the same or closest-matching product on JD, Pinduoduo, and Vipshop, then turn that evidence into a purchase decision against the Taobao baseline when available.\n\n## Common Reminders\n\n- Prefer visible desktop web pages.\n- Use exact product names first, then a lightly simplified query if results are sparse.\n- Compare like-for-like products only.\n- Keep the evidence chain simple: search page -> listing -> visible title/price/spec.\n- Do not use APIs, hidden JSON endpoints, or scraping shortcuts outside the browser tool.\n- Price is not enough; capture trust, conditions, and caveats.\n- The final answer should help the user choose, not just browse.\n- Read `category-playbooks.md` when the product category changes how much price should matter.\n\n## Baseline Discipline\n\n- Treat Taobao as the baseline only when a Taobao price, URL, screenshot, or user-provided listing detail exists.\n- If the user provides only a Taobao title, search against that identity but write `淘宝基准价未提供`.\n- Do not claim the user should switch away from Taobao unless Taobao's visible or user-provided price is known.\n- When the user provides the Taobao price, label it as user-provided unless verified in browser.\n\n## Matching Score Reminders\n\nUse the 100-point score from `SKILL.md`:\n- brand: 25\n- model / line / generation: 25\n- variant and must-match specs: 20\n- size / weight / count / packaging: 15\n- seller or channel trust: 10\n- evidence completeness: 5\n\nInterpretation:\n- 85-100: high, can drive recommendation\n- 65-84: medium, caveated recommendation only\n- below 65: low, reference/near-match only\n\nDowngrade to low when the result has a critical mismatch such as different generation, capacity, count, refurbished status, unclear warranty, trial size, or unverified authenticity in a risk-sensitive category.\n\n## Price Normalization Reminders\n\nBefore ranking, separate:\n- headline/list price\n- visible final or coupon-after price\n- coupon/member/subsidy/group-buy/payment/region conditions\n- unit basis such as per pack, per sheet, per 100g, per ml, per pair, or per item\n- shipping, free-shipping threshold, installation, warranty, invoice, or delivery constraints\n\nUse conservative comparison:\n- Prefer unconditional visible price when promo eligibility is unclear.\n- Treat member-only, group-buy, payment, and regional subsidy prices as conditional.\n- For multi-pack daily goods, compare unit price and total quantity.\n- For electronics, beauty, apparel, infant goods, medical/safety goods, and branded products, keep authenticity, warranty, and channel trust ahead of headline price.\n\n## Category Shortcut\n\nUse `references/category-playbooks.md` to decide which risk dominates:\n- warranty and invoice for electronics/appliances\n- authenticity, batch, and expiry for beauty/personal care\n- unit price, pack count, and freshness for food/daily goods\n- size, color, style code, and return policy for apparel/shoes/bags\n- official authorization and safety evidence for baby/health/safety products\n- ISBN/model/count and shipping for books/stationery/low-risk standard goods\n\nWhen category risk and headline price disagree, mention the conflict and let category risk control recommendation strength.\n\n## JD\n\n- Usually supports standard web search and listing pages well.\n- Prefer self-operated or official flagship listings when multiple near-identical results exist.\n- Watch for coupon text, plus/member pricing, trade-in, bank/payment offers, and promotional banners.\n- Distinguish list price from final promo price if both are shown.\n- Treat 京东自营 and official flagship stores as stronger trust signals when prices are close.\n- For electronics, appliances, beauty, infant goods, and regulated products, do not recommend a weaker channel solely because it is cheaper.\n\n## Pinduoduo\n\n- Web results may be noisier than JD.\n- Watch for subsidy labels, group-buy wording, coupon claims, and strong promo framing.\n- Matching confidence should be reduced when the seller, spec, or packaging is unclear.\n- If the browser experience is limited, use a public search engine with site restriction, then open the visible result page.\n- If the low price depends on 拼团 or 百亿补贴, call that out explicitly in the final recommendation.\n- Reduce recommendation strength when seller trust is unclear, even if the price looks excellent.\n- A very low PDD price should be treated as a lead to verify, not as a conclusion, unless match and seller evidence are strong.\n\n## Vipshop\n\n- Results may lean toward branded discount inventory and variant-specific listings.\n- Pay attention to size/color/version because discount channels often surface adjacent variants.\n- If the exact match is missing, mark the result as `近似款` instead of treating it as the same SKU.\n- Vipshop often wins on branded discounts but loses on exact spec coverage; do not overstate comparability.\n- Treat limited stock, size gaps, color mismatch, and flash-sale timing as decision caveats.\n\n## Evidence Standard\n\nFor each platform, aim to capture:\n- title\n- displayed price\n- visible final price or promo condition\n- variant/specification\n- seller/store if visible\n- URL\n- matching score and reason\n- a short matching note\n\nWhen visible, also capture:\n- official / self-operated / flagship status\n- coupon or membership dependency\n- subsidy / group-buy dependency\n- shipping promise\n- return / warranty / invoice / authenticity guarantee\n\nEvidence quality gate:\n- A listing should have title, price, spec/version, and URL before it can support a strong recommendation.\n- If two or more of those are missing, downgrade confidence to low.\n- Low-confidence candidates can be included, but they should not determine the final answer alone.\n\n## Output Discipline\n\nUse short, decision-friendly language.\n\nRecommended summary fields:\n- Taobao baseline status\n- category strategy and buying posture\n- lowest visible raw price\n- lowest visible comparable price\n- whether the comparison is apples-to-apples\n- major caveats: variant mismatch, coupon dependency, seller difference, shipping difference, warranty/invoice difference, regional limitation\n- recommended platform\n- recommendation strength: strong / weak / reference only / cannot judge\n- one-sentence reason the user should or should not switch away from Taobao\n- final checks before buying: final payable price, address-based delivery, coupon eligibility, stock, warranty, and return policy\n\nFile v2.1.0:skill-card.md\n\n## Description: <br>\nTaobao Price Compare helps compare a Taobao product against JD, PDD, and Vipshop using browser-visible evidence, normalized prices, same-item confidence scoring, and purchase guidance without login, order submission, or payment. <br>\n\nThis skill is ready for commercial/non-commercial use. <br>\n\n## Publisher: <br>\n[harrylabsj](https://clawhub.ai/user/harrylabsj) <br>\n\n### License/Terms of Use: <br>\nMIT-0 <br>\n\n\n## Use Case: <br>\nExternal shoppers and shopping assistants use this skill to compare a Taobao product against JD, Pinduoduo, and Vipshop, normalize visible prices and variants, and decide whether to buy, wait, stay with Taobao, or avoid. <br>\n\n### Deployment Geography for Use: <br>\nGlobal <br>\n\n## Known Risks and Mitigations: <br>\nRisk: Shopping comparisons may influence buying decisions using browser-visible prices and marketplace terms that can vary by account, address, coupon eligibility, stock, or time. <br>\nMitigation: Treat the output as decision support and personally recheck final price, coupons, delivery, returns, warranty, and stock before purchasing. <br>\nRisk: Marketplace browsing can lead to account, checkout, or payment flows. <br>\nMitigation: Use an isolated browser when possible, avoid credentials and payment details, and do not allow the skill to log in, submit orders, or initiate payment. <br>\n\n\n## Reference(s): <br>\n- [Category Playbooks](references/category-playbooks.md) <br>\n- [Site Notes](references/site-notes.md) <br>\n\n\n## Skill Output: <br>\n**Output Type(s):** [text, markdown, guidance] <br>\n**Output Format:** [Markdown with a decision summary, comparison table, match scores, risk notes, and final purchase checks] <br>\n**Output Parameters:** [1D] <br>\n**Other Properties Related to Output:** [Uses browser-visible marketplace evidence and asks the user to recheck final payable price, coupon eligibility, delivery, stock, returns, and warranty before purchasing.] <br>\n\n## Skill Version(s): <br>\n2.1.0 (source: ClawHub release evidence) <br>\n\n## Ethical Considerations: <br>\nUsers should evaluate whether this skill is appropriate for their environment, review any generated or modified files before relying on them, and apply their organization's safety, security, and compliance requirements before deployment. <br>\n\nArchive v1.2.0: 6 files, 18727 bytes\n\nFiles: README.md (3911b), references/category-playbooks.md (7975b), references/site-notes.md (6507b), skill-card.md (2661b), SKILL.md (20116b), _meta.json (145b)\n\nFile v1.2.0:SKILL.md\n\n---\nname: taobao-competitor-analyzer\ndescription: Compare a Taobao product with JD.com, Pinduoduo, and Vipshop using browser-visible evidence only, normalize prices, score same-item confidence, apply category-specific buying playbooks, and tell the user where it is actually worth buying. Use when the user wants to 查淘宝同款、比价、看竞品、分析值不值得买、判断哪个平台更划算、比较京东拼多多唯品会价格、找最低可见可比价、识别近似款风险, or get a cross-platform price check, buying recommendation, seller-risk comparison, or marketplace research report without using APIs.\n---\n\n# Taobao Competitor Analyzer\n\nCompare a Taobao product with the same or closest-matching listings on 京东、拼多多、唯品会 using the browser tool only. Work from a product name, Taobao title, link, screenshot, or user-provided baseline details. Return a compact purchase decision with visible price evidence, normalized comparability, match score, category-aware platform fit, recommendation strength, and risk notes.\n\nWhat makes this skill useful:\n- It compares comparable items instead of chasing misleading low prices.\n- It separates Taobao baseline price from competitor prices.\n- It normalizes price by spec, quantity, shipping, and promo conditions before ranking.\n- It scores same-item confidence before making a buying recommendation.\n- It applies category-specific risk thresholds instead of treating shoes, phones, skincare, and paper towels the same way.\n- It explains whether a lower price depends on coupons, membership, subsidy, or group-buy.\n- It ends with a clear buy / wait / avoid recommendation instead of just a table.\n\n## When To Use\n\nUse this skill when the user is effectively asking:\n- 这件淘宝商品别的平台多少钱\n- 有没有同款或更划算的平台\n- 京东 / 拼多多 / 唯品会 哪个更值得买\n- 这几个平台价格差这么多正常吗\n- 帮我做一个同款比价和购买建议\n- 这个淘宝商品换平台买值不值\n- 同样价差下哪个平台风险更低\n\nThe skill should optimize for purchase decisions, not raw data collection.\n\n## Commerce Matrix\n\nThis skill is the cross-platform comparison node in the shopping matrix.\n\nPrefer nearby skills when the task is narrower:\n- `taobao-shopping` for Taobao-only listing and seller evaluation\n- `jd-shopping` for trust-first self-operated buying\n- `pdd-shopping` for low-price and subsidy-first buying\n- `tianmao` for flagship-store and authenticity-first buying\n- `vip` for branded discount and flash-sale buying\n- `alibaba-shopping` when the user first needs to choose between Taobao, Tmall, and 1688\n\n## Workflow\n\n1. Normalize the Taobao baseline.\n   - Extract product identity: brand, model / series, variant, size / spec / count, color / flavor / version, packaging, and must-match attributes.\n   - Record the Taobao price only if the user provides it or it is browser-visible.\n   - If no Taobao price is available, write `淘宝基准价未提供` and do not imply whether switching platforms saves money versus Taobao.\n2. Classify the product category and buying posture.\n   - Pick the closest category from `references/category-playbooks.md`.\n   - Infer whether the user is price-first, trust-first, speed-first, exact-variant-first, or research-first.\n   - If the category or posture would change the recommendation and cannot be inferred, ask one short follow-up.\n3. If the input is only a Taobao-style long title, compress it into the smallest searchable core:\n   - brand\n   - model / series\n   - size / spec / count\n   - key variant\n4. Search the exact or lightly simplified keyword on:\n   - 京东\n   - 拼多多\n   - 唯品会\n5. Stay in browser-driven flows only. Do not call site APIs, hidden JSON endpoints, app-only interfaces, or unofficial scrapers.\n6. Extract the top relevant visible results from each site.\n7. Normalize each visible price into a comparable basis before ranking.\n8. Score same-item comparability before judging price.\n9. Apply category playbook gates before recommending a lower-price platform.\n10. Decide with recommendation strength: `强推荐`, `弱推荐`, `仅供参考`, or `无法判断`.\n11. End with a concrete recommendation: buy on which platform, stay with Taobao, wait, or avoid for now.\n\n## Input Rules\n\nRequire a product identity as input.\n\nPrefer one of these inputs:\n- Taobao product link, screenshot, title, or visible listing details\n- precise product name\n- brand + model + spec\n- product name plus intended use, if there are multiple variants\n- Taobao visible price or expected Taobao budget, if the user wants a switch / stay decision\n- category cue, such as phone, skincare, baby formula, shoes, snacks, tissue, appliance, or supplement\n- buying priority, such as lowest price, official/authenticity, fast delivery, easy returns, warranty/invoice, exact color/size, or competitor research\n\nIf the product name is too broad, ask one short follow-up to narrow it, for example:\n- brand\n- model\n- size/specification\n- package count\n- flavor/color/version\n- whether the user prioritizes lowest price, authenticity/after-sales, or delivery speed\n\nGood inputs:\n- `Apple AirPods Pro 2`\n- `维达抽纸 3层 100抽 24包`\n- `耐克 Air Zoom Pegasus 41 男款`\n- `这个淘宝价 129，帮我看看京东拼多多唯品会有没有更值的同款`\n\nWeak inputs that need clarification:\n- `纸巾`\n- `耳机`\n- `运动鞋`\n\nIf the user pastes a very long Taobao title, do not ask them to rewrite it unless it is truly ambiguous. You should clean and normalize it yourself first.\n\n## Taobao Baseline Rules\n\nTreat Taobao as the baseline only when baseline evidence exists.\n\nCapture when visible or user-provided:\n- Taobao title\n- Taobao displayed price\n- coupon-after, membership, or promo wording\n- spec / version / count / packaging\n- seller/store type\n- shipping or delivery note\n- URL or screenshot context\n\nIf the baseline is only a title:\n- compare competitor platforms against the title identity\n- say `淘宝基准价未提供，本次只比较竞品平台可见结果`\n- avoid saying `值得从淘宝换平台` unless the user later provides Taobao price or visible Taobao evidence\n\nIf the user provides a Taobao price:\n- call it `用户提供的淘宝基准价`\n- compare it to normalized competitor prices\n- flag that final payable price may change with address, coupon eligibility, account status, and stock\n\n## Browser Execution Rules\n\n- Prefer the isolated OpenClaw browser unless the user explicitly asks to use their Chrome tab.\n- Start with one tab per site when practical.\n- Re-snapshot after navigation or major DOM changes.\n- If a site shows login walls, anti-bot interstitials, region prompts, or app-download overlays, use the visible web result if possible and mention the limitation.\n- If a site blocks access completely, report it instead of trying to bypass it.\n- Do not fabricate missing prices.\n- Prefer visible search/listing pages over deep product pages when one platform is unstable.\n- Capture enough evidence to justify the recommendation, not just enough to fill a table.\n\n## Search Targets\n\nUse the standard web search pages when possible:\n\n- 京东: search for the product name on jd.com\n- 拼多多: search for the product name on pinduoduo.com or the visible web listing/search experience available in browser\n- 唯品会: search for the product name on vip.com\n\nIf direct site search is unstable in browser, use a public search engine query constrained to the site, then open the most relevant visible result. Example pattern:\n- `site:jd.com 商品名`\n- `site:pinduoduo.com 商品名`\n- `site:vip.com 商品名`\n\nStill use browser navigation for the actual evidence collection.\n\n## Category And Preference Routing\n\nRead `references/category-playbooks.md` when the product category is obvious or when price, trust, warranty, freshness, sizing, or authenticity can change the decision.\n\nClassify the category before final recommendation:\n- `数码/家电`: model, version, warranty, invoice, self-operated/official channel, installation, trade-in, refurbished/open-box risk.\n- `美妆/个护`: official channel, batch/expiry, sample/trial size, authenticity, sealed packaging, return limits.\n- `食品/日用品`: unit price, pack count, shelf life, shipping threshold, heavy-item delivery, commodity substitutability.\n- `服饰/鞋包`: exact color/size/style code, season/version, authenticity, return convenience, stock by size.\n- `母婴/健康/安全`: official/authorized channel, registration/approval when relevant, expiry, warranty, return policy, safety risk.\n- `图书/文具/低风险标品`: ISBN/model/count, shipping, bundle contents, seller reliability, unit price.\n- `不确定/混合`: say which playbook you used and why.\n\nInfer buying posture:\n- `低价优先`: maximize normalized savings, but only after match and risk gates pass.\n- `正品/售后优先`: prefer official, self-operated, flagship, authorized, invoice, warranty, and return clarity.\n- `速度优先`: prefer visible delivery promise and stable fulfillment even if not the cheapest.\n- `精确款优先`: prioritize exact variant, size, color, model, batch, or package count.\n- `研究优先`: broaden to near matches and substitutes, but clearly separate them from exact matches.\n\nUse category-specific price-delta thresholds:\n- For high-risk or warranty-heavy goods, require a meaningful saving before recommending a weaker channel.\n- For low-risk commodities, a clear normalized unit-price win can justify switching more easily.\n- If the category playbook conflicts with the lowest visible price, the playbook wins.\n\n## Matching Score\n\nTreat listings as comparable only when the core attributes align.\n\nScore each candidate out of 100:\n- Brand: 25\n- Product line / model / generation: 25\n- Variant and must-match specs: 20\n- Size / weight / count / packaging: 15\n- Seller/channel trust when relevant: 10\n- Evidence completeness: 5\n\nUse these match bands:\n- `高` / 85-100: same item or same SKU-equivalent listing; price comparison can drive the recommendation.\n- `中` / 65-84: near match with one explainable difference; include it with caveats, and make only a weak recommendation unless the tradeoff is obvious.\n- `低` / below 65: reference or substitute only; do not call it cheaper than Taobao or cheaper than another exact match.\n\nForce the match band to `低` when any critical mismatch appears:\n- different model, generation, storage, capacity, size, flavor, color, or pack count\n- refurbished / open-box / parallel import when the baseline is standard new domestic stock\n- trial/sample size compared with regular size\n- unclear warranty or authenticity for safety-sensitive or high-counterfeit-risk categories\n- missing enough information to confirm the listing is the same item\n\nApply category overrides:\n- For high-counterfeit, safety-sensitive, or warranty-heavy categories, reduce the score when official/self-operated/authorized evidence is missing.\n- For low-risk commodities, do not over-penalize seller/channel differences when the spec, count, and unit basis are clear.\n- For apparel and shoes, exact size, color, style code, and return policy can matter as much as headline model name.\n- For beauty, baby, food, and supplements, expiry, batch, sealed packaging, and authenticity cues can be critical match evidence.\n\nIf there is no close match on a platform:\n- say `未找到足够接近的同款`\n- optionally include the nearest visible alternative, clearly labeled as `近似款` or `替代款`\n\n## Price Normalization Rules\n\nDo not rank prices until they are normalized.\n\nFor each candidate, separate these fields when visible:\n- `标价`: the headline displayed price\n- `可见到手价`: coupon-after, subsidy, member, or promo price, only when the condition is visible\n- `优惠条件`: coupon, membership, group-buy, payment method, limited-time sale, regional subsidy, or account qualification\n- `单位基准`: per item, per pack, per 100g, per ml, per sheet, per pair, or another meaningful unit\n- `运费/门槛`: shipping fee, free-shipping threshold, installation fee, or delivery limitation\n\nUse the most conservative comparable price:\n- Prefer unconditional visible price when coupon eligibility is unclear.\n- Use coupon-after price only when the page makes the condition visible and likely attainable.\n- Keep member-only, group-buy, subsidy, bank-card, and region-dependent prices separate from normal prices.\n- For multi-pack goods, compare unit price and total package value, not only headline price.\n- For electronics, apparel, beauty, and branded goods, do not let a lower price override warranty, channel, authenticity, or version mismatch.\n- Use the category playbook's preferred unit basis when available.\n\nNever recommend a cheaper platform when the cheaper listing is only cheaper because of:\n- lower specification\n- different quantity or packaging\n- unclear seller trust\n- member-only or coupon-after price not available to most users\n- group-buy requirement the user may not want\n- missing shipping, installation, warranty, or invoice cost\n\n## Decision Rules\n\nUse this order of judgment:\n\n1. Is the candidate the same item or only a near match\n2. Is the visible price directly comparable after normalization\n3. Does the category playbook allow price to dominate, or should trust/warranty/authenticity/freshness/sizing dominate\n4. Is the seller/store trust level similar\n5. Are coupon, membership, subsidy, or group-buy conditions required\n6. Are shipping speed, warranty, invoice, return rights, and after-sales meaningfully different\n7. Does the user's likely preference favor lowest price, trust, delivery speed, exact variant availability, or broad research\n\nRecommendation strength:\n- `强推荐`: high match, clear normalized price, trustworthy seller/channel, and the advantage is material.\n- `弱推荐`: high or medium match, but with one meaningful caveat such as coupon dependency, seller difference, delivery uncertainty, category risk, or missing Taobao baseline price.\n- `仅供参考`: partial evidence, noisy results, medium/low match, or blocked platform pages.\n- `无法判断`: no comparable listings, missing baseline for the user question, or evidence too incomplete to support a purchase decision.\n\nDo not use one universal price threshold:\n- A 5-8% saving may be weak for a phone if warranty/channel is worse.\n- A 10-20% normalized unit-price saving may matter for low-risk daily goods.\n- Any saving can be irrelevant if the exact size, color, model, expiry, or warranty does not match.\n\nIf Taobao is not actually the best option, say so directly. If none of the results are truly comparable, explicitly say `当前不适合做强结论`.\n\n## What To Capture Per Site\n\nCapture only information visible on the page. Prefer the first 1-3 relevant results.\n\nFor each selected listing, collect when visible:\n- platform\n- title\n- displayed price\n- visible final/promo price and condition\n- unit price or normalized price basis, when applicable\n- package/specification\n- store/seller name\n- delivery or shipping note\n- URL\n- match score and band\n- category and buying-posture note when it changes the conclusion\n- confidence: high / medium / low\n- note about why it matches or why it is only approximate\n\nAlso capture, when visible and relevant:\n- official/self-operated/flagship indicator\n- coupon or subsidy dependency\n- group-buy requirement\n- delivery speed or shipping promise\n- return, warranty, invoice, or authenticity guarantee\n\nEvidence quality gate:\n- A candidate should include at least title, displayed price, spec/version, and URL to drive a recommendation.\n- If two or more of those fields are missing, downgrade confidence to low.\n- Low-confidence candidates can appear in the table but cannot be the sole basis for `强推荐`.\n\n## Output Format\n\nReturn a decision first, then the evidence table.\n\nStart with a short verdict block:\n\n- `淘宝基准`: visible/user-provided price, or `淘宝基准价未提供`\n- `品类策略`: category playbook used and why\n- `购买偏好`: inferred or user-provided priority\n- `推荐平台`: platform or `暂不建议换平台`\n- `推荐强度`: `强推荐` / `弱推荐` / `仅供参考` / `无法判断`\n- `最低可见可比价`: platform + normalized price basis, only among comparable items\n- `值不值得换平台`: yes / no / depends, with one sentence\n- `主要原因`\n- `风险点`\n- `核验时间`: include date/time or say browser-visible at time of checking\n\nThen return a concise comparison table and short notes.\n\nUse a table like this:\n\n| 平台 | 商品标题 | 标价 | 可见到手价/条件 | 单位价/基准 | 规格/版本 | 店铺 | 匹配分 | 匹配度 | 备注 |\n|---|---|---:|---|---|---|---|---:|---|---|\n| 淘宝 | ... | ¥... | ... | ... | ... | ... | 基准 | 基准 | 用户提供/页面可见 |\n| 京东 | ... | ¥... | ... | ... | ... | ... | 92 | 高 | 同品牌同规格 |\n| 拼多多 | ... | ¥... | ... | ... | ... | ... | 78 | 中 | 规格接近，包装不同 |\n| 唯品会 | ... | ¥... | ... | ... | ... | ... | 52 | 低 | 仅找到近似款 |\n\nThen add:\n- `最低可见价`: platform and raw visible price\n- `最低可比价`: platform and normalized comparable price\n- `可比性判断`: high / medium / low, with reason\n- `品类判断`: why this category should prioritize price, trust, delivery, warranty, freshness, authenticity, or exact variant\n- `风险提示`: differences in package, seller, promo timing, membership price, shipping, warranty, invoice, or coupon requirements\n- `购买建议`: 直接买 / 可等等 / 只建议在某平台买 / 暂不建议下单\n- `下单前核对`: final payable price, address-based delivery, coupon eligibility, stock, warranty, and return policy\n\n## Interpretation Rules\n\n- Do not claim a platform is cheaper unless the compared items are materially comparable.\n- Separate `标价` from coupon-after price when the page makes that distinction.\n- Mention when a price may depend on membership, flash sale, subsidy, payment method, group-buy, or region.\n- If search results are noisy, prefer accuracy over completeness.\n- If the Taobao baseline price is missing, do not say the user should switch away from Taobao; say which competitor has the best visible comparable offer.\n- If the Taobao baseline is not actually the best option, say so directly.\n- If category-specific risk outweighs the price difference, recommend staying with the safer channel or waiting.\n- If none of the results are truly comparable, explicitly say `当前不适合做强结论`.\n- If the user seems purchase-ready, optimize the answer for actionability: where to buy, what to verify before paying, and what tradeoff they are accepting.\n\n## Example Requests\n\n- `帮我查一下“德芙黑巧克力 84g”在京东、拼多多、唯品会的价格`\n- `对比一下“iPhone 16 Pro 256GB”在几个平台上的可见报价`\n- `把这个淘宝商品名拿去京东、拼多多、唯品会搜同款，做个价格表`\n- `这个淘宝商品有没有更便宜但靠谱的平台`\n- `帮我判断这件商品有没有必要从淘宝换到京东买`\n- `淘宝价 129，这个商品换平台买值不值`\n- `别只比价，也告诉我哪个平台更值得下单`\n\n## Failure Handling\n\nIf one or more sites cannot be accessed or searched reliably, still return a partial result and list:\n- which site failed\n- what was attempted\n- whether the failure was due to login wall, anti-bot page, timeout, app-only flow, or missing web search results\n\nIf evidence is partial, downgrade recommendation strength.\n\nDo not fill missing evidence with guesses. Use:\n- `未见明确价格`\n- `未见规格`\n- `未见店铺信息`\n- `无法确认同款`\n\n## Resource\n\n- Read `references/site-notes.md` when you need execution reminders for JD, Pinduoduo, and Vipshop search behavior, evidence standards, normalized pricing, and recommendation gates.\n- Read `references/category-playbooks.md` when category, user preference, authenticity, warranty, freshness, sizing, safety, or platform fit can change the recommendation.\n\nFile v1.2.0:README.md\n\n# Taobao Competitor Analyzer\n\nCross-platform shopping decision skill for Taobao users.\n\nThis skill checks the same or closest-matching product on:\n- JD.com\n- Pinduoduo\n- Vipshop\n\nThen it answers the question users actually care about:\n\n`Should I keep buying this on Taobao, switch platforms, wait, or avoid this deal?`\n\n## What It Does\n\n- Compares visible prices across major Chinese marketplaces\n- Uses Taobao as the baseline when a Taobao price, link, screenshot, or visible listing is available\n- Matches brand, model, spec, count, and packaging before comparing price\n- Scores same-item confidence so near matches do not masquerade as cheaper exact matches\n- Normalizes headline price, coupon-after price, member price, subsidy price, group-buy price, shipping, and unit price\n- Applies category playbooks so phones, skincare, baby goods, shoes, snacks, and paper towels are judged differently\n- Flags near-match risk instead of pretending similar items are identical\n- Weighs seller trust, shipping, warranty, invoice, and after-sales guarantees\n- Returns a recommendation strength: strong, weak, reference-only, or cannot judge\n\n## v1.1.0 Highlights\n\n- Taobao baseline handling: no baseline price means no unsupported \"switch away from Taobao\" claim\n- 100-point match scoring across brand, model, specs, quantity, channel trust, and evidence completeness\n- Price normalization for unit price, coupon dependency, membership, subsidy, group-buy, and shipping caveats\n- Evidence quality gate for title, price, spec/version, and URL before a listing can drive a strong recommendation\n- Expanded verdict block with recommendation strength, lowest comparable price, risks, and final checks before purchase\n\n## v1.2.0 Highlights\n\n- Category-aware playbooks for electronics/appliances, beauty/personal care, food/daily goods, apparel/shoes/bags, baby/health/safety, and low-risk standard goods\n- Buying-posture routing for lowest price, authenticity/after-sales, delivery speed, exact variant, and research-first workflows\n- Platform-fit judgment so JD, PDD, and Vipshop are weighed differently by category instead of by headline price alone\n- Category-specific price-delta thresholds: low-risk commodities can switch on unit-price wins, while warranty/authenticity categories need stronger evidence\n- New `品类策略` and `购买偏好` fields in the verdict block\n\n## Best Use Cases\n\n- Compare a Taobao listing with JD / PDD / Vipshop\n- Check whether a \"cheaper\" result is actually the same SKU\n- Decide if it is worth switching platforms\n- Do fast competitor research for consumer products\n- Find the lowest visible comparable price with caveats\n- Separate normal price from coupon, member, subsidy, and group-buy conditions\n- Decide whether a low price is worth the category-specific risk\n\n## Example Prompts\n\n- `帮我查这个淘宝商品在京东、拼多多、唯品会有没有同款`\n- `淘宝价 129，这个商品换到京东买值不值`\n- `这款护肤品拼多多便宜很多，风险大不大`\n- `这双鞋唯品会有近似款，能不能买`\n- `别只比价，告诉我哪个平台更值得买`\n- `Compare this Taobao product with JD, PDD, and Vipshop and recommend where to buy`\n\n## Output Style\n\nThe skill is optimized to return:\n- Taobao baseline status\n- recommended platform\n- recommendation strength\n- lowest visible comparable price\n- normalized unit/condition basis\n- category strategy and inferred buying priority\n- whether the comparison is apples-to-apples\n- major risks and caveats\n- a direct buy / wait / avoid recommendation\n\n## Positioning\n\nThis is not a generic shopping browser helper.\n\nIt is a focused purchase-decision skill for users who want:\n- faster price comparison\n- fewer fake \"cheap\" matches\n- clearer marketplace tradeoff analysis\n- conservative recommendations when evidence is incomplete\n- category-aware buying advice rather than one-size-fits-all price sorting\n\nFile v1.2.0:_meta.json\n\n{\n  \"ownerId\": \"kn77zzg9p845zanvy6vrf76k7d81mcnm\",\n  \"slug\": \"taobao-competitor-analyzer\",\n  \"version\": \"1.2.0\",\n  \"publishedAt\": 1780838902529\n}\n\nFile v1.2.0:references/category-playbooks.md\n\n# Category Playbooks\n\n## Goal\n\nUse these playbooks after normalizing the Taobao baseline and before making the final recommendation. The goal is to avoid one-size-fits-all price sorting. A lower price means different things for phones, skincare, baby goods, shoes, snacks, and paper towels.\n\n## How To Use\n\n1. Classify the product into the closest category.\n2. Infer the user's buying posture: low price, trust/after-sales, delivery speed, exact variant, or research.\n3. Apply the category's must-check attributes and risk gates.\n4. Decide whether the price gap is large enough to overcome category risk.\n5. State the playbook in the verdict as `品类策略`.\n\nIf the category is unclear, write the assumption and keep recommendation strength at `弱推荐` or lower.\n\n## Buying Postures\n\n### Low Price First\n\n- Use normalized unit price after matching gates pass.\n- Prefer unconditional visible prices over conditional coupon/member/group-buy prices.\n- Strong recommendations still need high match and enough seller/channel evidence.\n\n### Trust And After-Sales First\n\n- Prefer official, self-operated, flagship, or authorized channels.\n- Treat invoice, warranty, returns, and authenticity evidence as decision drivers.\n- A cheaper weak-channel listing usually becomes `弱推荐` or `仅供参考`.\n\n### Speed First\n\n- Prefer visible delivery promise, local stock, self-operated logistics, and stable fulfillment.\n- Downgrade group-buy, pre-sale, limited stock, and unclear shipping.\n- If delivery is address-dependent and not visible, include it in `下单前核对`.\n\n### Exact Variant First\n\n- Prioritize exact model, color, size, flavor, version, pack count, batch, and bundle contents.\n- Do not treat near matches as cheaper exact matches.\n- For apparel, shoes, beauty, and electronics, variant mismatch usually blocks `强推荐`.\n\n### Research First\n\n- Include exact matches, near matches, and substitutes in separate groups.\n- Do not rank substitutes against exact matches as if they are identical.\n- Use this mode for marketplace research, category scans, and competitor mapping.\n\n## Electronics And Appliances\n\nExamples: phones, headphones, laptops, cameras, routers, appliances, smart devices.\n\nMust check:\n- exact model, generation, storage/capacity, color, region/version, bundle, and warranty\n- new vs refurbished/open-box/parallel import\n- official/self-operated/authorized channel\n- invoice, warranty, installation, return policy, trade-in, and delivery timing\n\nPlatform fit:\n- JD is often stronger when self-operated, official, invoice, warranty, fast delivery, or installation matters.\n- PDD can be considered when subsidy/official evidence is visible and the SKU is exact.\n- Vipshop is usually weaker for exact electronics coverage unless the listing is clearly official and exact.\n\nRecommendation gate:\n- Do not recommend a weaker channel for small savings.\n- Require high match and clear warranty/channel evidence for `强推荐`.\n- If savings are modest and JD/Taobao official support is stronger, recommend staying with the safer channel or waiting.\n\n## Beauty And Personal Care\n\nExamples: skincare, perfume, makeup, haircare, oral care, personal care devices.\n\nMust check:\n- exact product line, volume, shade, scent, set contents, version, and packaging\n- official/flagship/authorized seller cues\n- batch, expiry, sealed packaging, import/domestic version, sample/trial size\n- return limits and authenticity guarantee\n\nPlatform fit:\n- Tmall/Taobao flagship, JD official/self-operated, and brand official channels are stronger trust signals.\n- Vipshop can be attractive for branded discount inventory when exact variant and channel confidence are visible.\n- PDD low prices need extra caution unless official/subsidy/channel evidence is clear.\n\nRecommendation gate:\n- Do not let a low price override unclear authenticity, sample size, expiry, or seller trust.\n- For branded skincare and perfume, weak-channel savings should usually be `仅供参考`.\n- If exact shade/volume/set differs, mark as near match or substitute.\n\n## Food And Daily Goods\n\nExamples: snacks, drinks, rice/oil, tissue, detergent, pet food, household consumables.\n\nMust check:\n- unit count, weight, volume, flavor, pack count, expiration/shelf life, and packaging\n- unit price basis such as per 100g, per bottle, per pack, per sheet, or per item\n- shipping fee, free-shipping threshold, heavy-item delivery, and regional stock\n- brand and seller reliability for food, pet food, and ingestible items\n\nPlatform fit:\n- PDD can win on low-risk commodities when spec, count, and unit basis are clear.\n- JD can win when speed, heavy-item delivery, after-sales, or self-operated reliability matters.\n- Vipshop is less central unless it has exact branded inventory or bundle discounts.\n\nRecommendation gate:\n- A clear normalized unit-price win can justify switching for low-risk daily goods.\n- For food, pet food, or anything ingested, expiry and seller reliability still matter.\n- Do not compare different pack counts without unit price.\n\n## Apparel, Shoes, And Bags\n\nExamples: clothing, sneakers, sports shoes, bags, accessories.\n\nMust check:\n- brand, style code/model, size, color, gender, season/version, material, and bundle/accessories\n- official/authorized seller when counterfeiting risk is meaningful\n- return/exchange policy, size availability, stock, and authenticity guarantee\n\nPlatform fit:\n- Vipshop can be strong for branded discount inventory when size/color/style code match exactly.\n- JD/Tmall/official channels are stronger when authenticity and returns matter.\n- PDD and generic Taobao listings require caution for branded goods unless evidence is strong.\n\nRecommendation gate:\n- Size/color mismatch usually means near match, not same item.\n- If returns are unclear for apparel/shoes, downgrade recommendation strength.\n- Do not recommend a cheaper listing for branded shoes/bags without strong authenticity evidence.\n\n## Baby, Health, And Safety\n\nExamples: infant formula, baby products, supplements, medicine-adjacent goods, helmets, batteries, chargers, appliances with safety implications, medical devices.\n\nMust check:\n- official/authorized seller, registration/approval where relevant, batch, expiry, standard/certification, warranty, and return policy\n- exact model/spec and safety certification\n- do not provide medical claims or health guarantees\n\nPlatform fit:\n- Prefer official, self-operated, flagship, authorized, or brand channels.\n- Treat very low prices from unclear sellers as risk signals.\n- PDD or marketplace third-party listings need unusually strong evidence to be more than `仅供参考`.\n\nRecommendation gate:\n- Safety and authenticity override price.\n- Do not issue `强推荐` for unclear-channel baby/health/safety goods.\n- If evidence is weak, recommend official/self-operated channels or waiting.\n\n## Books, Stationery, And Low-Risk Standard Goods\n\nExamples: books, notebooks, pens, cables with low risk, small office supplies, simple accessories.\n\nMust check:\n- ISBN/model, edition, count, color, size, bundle contents, shipping, and seller reliability\n- for books, distinguish正版,影印,二手,预售,套装, and different editions\n\nPlatform fit:\n- PDD and Taobao can be reasonable for low-risk standard goods when listing details are clear.\n- JD can be better for fast shipping, invoices, and standardized fulfillment.\n- Vipshop is usually relevant only for branded stationery or limited discount inventory.\n\nRecommendation gate:\n- Price can matter more once edition/model/count are confirmed.\n- Do not compare different editions, bundle counts, or used/new condition as exact matches.\n\n## Unknown Or Mixed Category\n\nUse this when the item does not fit cleanly.\n\nRules:\n- State the assumed category.\n- Use the strictest relevant risk gate among possible categories.\n- Keep recommendation at `弱推荐` or lower unless evidence is very strong.\n- Ask one short follow-up if the category would materially change the purchase advice.\n\nFile v1.2.0:references/site-notes.md\n\n# Site Notes\n\n## Goal\n\nUse browser-visible marketplace pages to collect price evidence for the same or closest-matching product on JD, Pinduoduo, and Vipshop, then turn that evidence into a purchase decision against the Taobao baseline when available.\n\n## Common Reminders\n\n- Prefer visible desktop web pages.\n- Use exact product names first, then a lightly simplified query if results are sparse.\n- Compare like-for-like products only.\n- Keep the evidence chain simple: search page -> listing -> visible title/price/spec.\n- Do not use APIs, hidden JSON endpoints, or scraping shortcuts outside the browser tool.\n- Price is not enough; capture trust, conditions, and caveats.\n- The final answer should help the user choose, not just browse.\n- Read `category-playbooks.md` when the product category changes how much price should matter.\n\n## Baseline Discipline\n\n- Treat Taobao as the baseline only when a Taobao price, URL, screenshot, or user-provided listing detail exists.\n- If the user provides only a Taobao title, search against that identity but write `淘宝基准价未提供`.\n- Do not claim the user should switch away from Taobao unless Taobao's visible or user-provided price is known.\n- When the user provides the Taobao price, label it as user-provided unless verified in browser.\n\n## Matching Score Reminders\n\nUse the 100-point score from `SKILL.md`:\n- brand: 25\n- model / line / generation: 25\n- variant and must-match specs: 20\n- size / weight / count / packaging: 15\n- seller or channel trust: 10\n- evidence completeness: 5\n\nInterpretation:\n- 85-100: high, can drive recommendation\n- 65-84: medium, caveated recommendation only\n- below 65: low, reference/near-match only\n\nDowngrade to low when the result has a critical mismatch such as different generation, capacity, count, refurbished status, unclear warranty, trial size, or unverified authenticity in a risk-sensitive category.\n\n## Price Normalization Reminders\n\nBefore ranking, separate:\n- headline/list price\n- visible final or coupon-after price\n- coupon/member/subsidy/group-buy/payment/region conditions\n- unit basis such as per pack, per sheet, per 100g, per ml, per pair, or per item\n- shipping, free-shipping threshold, installation, warranty, invoice, or delivery constraints\n\nUse conservative comparison:\n- Prefer unconditional visible price when promo eligibility is unclear.\n- Treat member-only, group-buy, payment, and regional subsidy prices as conditional.\n- For multi-pack daily goods, compare unit price and total quantity.\n- For electronics, beauty, apparel, infant goods, medical/safety goods, and branded products, keep authenticity, warranty, and channel trust ahead of headline price.\n\n## Category Shortcut\n\nUse `references/category-playbooks.md` to decide which risk dominates:\n- warranty and invoice for electronics/appliances\n- authenticity, batch, and expiry for beauty/personal care\n- unit price, pack count, and freshness for food/daily goods\n- size, color, style code, and return policy for apparel/shoes/bags\n- official authorization and safety evidence for baby/health/safety products\n- ISBN/model/count and shipping for books/stationery/low-risk standard goods\n\nWhen category risk and headline price disagree, mention the conflict and let category risk control recommendation strength.\n\n## JD\n\n- Usually supports standard web search and listing pages well.\n- Prefer self-operated or official flagship listings when multiple near-identical results exist.\n- Watch for coupon text, plus/member pricing, trade-in, bank/payment offers, and promotional banners.\n- Distinguish list price from final promo price if both are shown.\n- Treat 京东自营 and official flagship stores as stronger trust signals when prices are close.\n- For electronics, appliances, beauty, infant goods, and regulated products, do not recommend a weaker channel solely because it is cheaper.\n\n## Pinduoduo\n\n- Web results may be noisier than JD.\n- Watch for subsidy labels, group-buy wording, coupon claims, and strong promo framing.\n- Matching confidence should be reduced when the seller, spec, or packaging is unclear.\n- If the browser experience is limited, use a public search engine with site restriction, then open the visible result page.\n- If the low price depends on 拼团 or 百亿补贴, call that out explicitly in the final recommendation.\n- Reduce recommendation strength when seller trust is unclear, even if the price looks excellent.\n- A very low PDD price should be treated as a lead to verify, not as a conclusion, unless match and seller evidence are strong.\n\n## Vipshop\n\n- Results may lean toward branded discount inventory and variant-specific listings.\n- Pay attention to size/color/version because discount channels often surface adjacent variants.\n- If the exact match is missing, mark the result as `近似款` instead of treating it as the same SKU.\n- Vipshop often wins on branded discounts but loses on exact spec coverage; do not overstate comparability.\n- Treat limited stock, size gaps, color mismatch, and flash-sale timing as decision caveats.\n\n## Evidence Standard\n\nFor each platform, aim to capture:\n- title\n- displayed price\n- visible final price or promo condition\n- variant/specification\n- seller/store if visible\n- URL\n- matching score and reason\n- a short matching note\n\nWhen visible, also capture:\n- official / self-operated / flagship status\n- coupon or membership dependency\n- subsidy / group-buy dependency\n- shipping promise\n- return / warranty / invoice / authenticity guarantee\n\nEvidence quality gate:\n- A listing should have title, price, spec/version, and URL before it can support a strong recommendation.\n- If two or more of those are missing, downgrade confidence to low.\n- Low-confidence candidates can be included, but they should not determine the final answer alone.\n\n## Output Discipline\n\nUse short, decision-friendly language.\n\nRecommended summary fields:\n- Taobao baseline status\n- category strategy and buying posture\n- lowest visible raw price\n- lowest visible comparable price\n- whether the comparison is apples-to-apples\n- major caveats: variant mismatch, coupon dependency, seller difference, shipping difference, warranty/invoice difference, regional limitation\n- recommended platform\n- recommendation strength: strong / weak / reference only / cannot judge\n- one-sentence reason the user should or should not switch away from Taobao\n- final checks before buying: final payable price, address-based delivery, coupon eligibility, stock, warranty, and return policy\n\nFile v1.2.0:skill-card.md\n\n## Description: <br>\nCompare a Taobao product with JD.com, Pinduoduo, and Vipshop using browser-visible evidence only, normalize prices, score same-item confidence, apply category-specific buying playbooks, and tell the user where it is actually worth buying. <br>\n\nThis skill is ready for commercial/non-commercial use. <br>\n\n## Publisher: <br>\n[harrylabsj](https://clawhub.ai/user/harrylabsj) <br>\n\n### License/Terms of Use: <br>\nMIT-0 <br>\n\n\n## Use Case: <br>\nExternal shoppers and marketplace researchers use this skill to compare Taobao products against JD.com, Pinduoduo, and Vipshop before deciding whether to buy, wait, avoid, or switch platforms. It is intended for purchase-decision support based on browser-visible listing evidence, normalized price comparisons, category-specific risk gates, and seller or promotion caveats. <br>\n\n### Deployment Geography for Use: <br>\nGlobal; primarily useful for Chinese marketplace shopping contexts. <br>\n\n## Known Risks and Mitigations: <br>\nRisk: Marketplace prices, coupons, membership offers, subsidies, stock, and delivery terms can change between comparison time and checkout. <br>\nMitigation: Verify the final payable price, coupon eligibility, delivery address effects, stock, and shipping terms before buying. <br>\nRisk: A lower-price listing may be a near match rather than the same SKU, or may have weaker seller authenticity, warranty, invoice, return, or after-sales evidence. <br>\nMitigation: Confirm brand, model, variant, count, packaging, seller authenticity, warranty, invoice, and return policy before relying on the recommendation. <br>\n\n\n## Reference(s): <br>\n- [Category Playbooks](artifact/references/category-playbooks.md) <br>\n- [Site Notes](artifact/references/site-notes.md) <br>\n- [ClawHub Skill Page](https://clawhub.ai/harrylabsj/taobao-competitor-analyzer) <br>\n- [Publisher Profile](https://clawhub.ai/user/harrylabsj) <br>\n\n\n## Skill Output: <br>\n**Output Type(s):** [Text, Markdown, Guidance] <br>\n**Output Format:** [Markdown verdict block, comparison table, risk notes, and final buying recommendation] <br>\n**Output Parameters:** [1D] <br>\n**Other Properties Related to Output:** [Uses browser-visible evidence only; outputs recommendation strength, normalized price basis, match confidence, and final checkout checks.] <br>\n\n## Skill Version(s): <br>\n1.2.0 (source: ClawHub release evidence) <br>\n\n## Ethical Considerations: <br>\nUsers should evaluate whether this skill is appropriate for their environment, review any generated or modified files before relying on them, and apply their organization's safety, security, and compliance requirements before deployment. <br>\n\nArchive v1.1.0: 5 files, 12821 bytes\n\nFiles: README.md (2836b), references/site-notes.md (5762b), skill-card.md (2597b), SKILL.md (15636b), _meta.json (145b)\n\nFile v1.1.0:SKILL.md\n\n---\nname: taobao-competitor-analyzer\ndescription: Compare a Taobao product with JD.com, Pinduoduo, and Vipshop using browser-visible evidence only, normalize prices, score same-item confidence, and tell the user where it is actually worth buying. Use when the user wants to 查淘宝同款、比价、看竞品、分析值不值得买、判断哪个平台更划算、比较京东拼多多唯品会价格、找最低可见可比价、识别近似款风险, or get a cross-platform price check, buying recommendation, seller-risk comparison, or marketplace research report without using APIs.\n---\n\n# Taobao Competitor Analyzer\n\nCompare a Taobao product with the same or closest-matching listings on 京东、拼多多、唯品会 using the browser tool only. Work from a product name, Taobao title, link, screenshot, or user-provided baseline details. Return a compact purchase decision with visible price evidence, normalized comparability, match score, recommendation strength, and risk notes.\n\nWhat makes this skill useful:\n- It compares comparable items instead of chasing misleading low prices.\n- It separates Taobao baseline price from competitor prices.\n- It normalizes price by spec, quantity, shipping, and promo conditions before ranking.\n- It scores same-item confidence before making a buying recommendation.\n- It explains whether a lower price depends on coupons, membership, subsidy, or group-buy.\n- It ends with a clear buy / wait / avoid recommendation instead of just a table.\n\n## When To Use\n\nUse this skill when the user is effectively asking:\n- 这件淘宝商品别的平台多少钱\n- 有没有同款或更划算的平台\n- 京东 / 拼多多 / 唯品会 哪个更值得买\n- 这几个平台价格差这么多正常吗\n- 帮我做一个同款比价和购买建议\n- 这个淘宝商品换平台买值不值\n\nThe skill should optimize for purchase decisions, not raw data collection.\n\n## Commerce Matrix\n\nThis skill is the cross-platform comparison node in the shopping matrix.\n\nPrefer nearby skills when the task is narrower:\n- `taobao-shopping` for Taobao-only listing and seller evaluation\n- `jd-shopping` for trust-first self-operated buying\n- `pdd-shopping` for low-price and subsidy-first buying\n- `tianmao` for flagship-store and authenticity-first buying\n- `vip` for branded discount and flash-sale buying\n- `alibaba-shopping` when the user first needs to choose between Taobao, Tmall, and 1688\n\n## Workflow\n\n1. Normalize the Taobao baseline.\n   - Extract product identity: brand, model / series, variant, size / spec / count, color / flavor / version, packaging, and must-match attributes.\n   - Record the Taobao price only if the user provides it or it is browser-visible.\n   - If no Taobao price is available, write `淘宝基准价未提供` and do not imply whether switching platforms saves money versus Taobao.\n2. If the input is only a Taobao-style long title, compress it into the smallest searchable core:\n   - brand\n   - model / series\n   - size / spec / count\n   - key variant\n3. Search the exact or lightly simplified keyword on:\n   - 京东\n   - 拼多多\n   - 唯品会\n4. Stay in browser-driven flows only. Do not call site APIs, hidden JSON endpoints, app-only interfaces, or unofficial scrapers.\n5. Extract the top relevant visible results from each site.\n6. Normalize each visible price into a comparable basis before ranking.\n7. Score same-item comparability before judging price.\n8. Decide with recommendation strength: `强推荐`, `弱推荐`, `仅供参考`, or `无法判断`.\n9. End with a concrete recommendation: buy on which platform, stay with Taobao, wait, or avoid for now.\n\n## Input Rules\n\nRequire a product identity as input.\n\nPrefer one of these inputs:\n- Taobao product link, screenshot, title, or visible listing details\n- precise product name\n- brand + model + spec\n- product name plus intended use, if there are multiple variants\n- Taobao visible price or expected Taobao budget, if the user wants a switch / stay decision\n\nIf the product name is too broad, ask one short follow-up to narrow it, for example:\n- brand\n- model\n- size/specification\n- package count\n- flavor/color/version\n- whether the user prioritizes lowest price, authenticity/after-sales, or delivery speed\n\nGood inputs:\n- `Apple AirPods Pro 2`\n- `维达抽纸 3层 100抽 24包`\n- `耐克 Air Zoom Pegasus 41 男款`\n- `这个淘宝价 129，帮我看看京东拼多多唯品会有没有更值的同款`\n\nWeak inputs that need clarification:\n- `纸巾`\n- `耳机`\n- `运动鞋`\n\nIf the user pastes a very long Taobao title, do not ask them to rewrite it unless it is truly ambiguous. You should clean and normalize it yourself first.\n\n## Taobao Baseline Rules\n\nTreat Taobao as the baseline only when baseline evidence exists.\n\nCapture when visible or user-provided:\n- Taobao title\n- Taobao displayed price\n- coupon-after, membership, or promo wording\n- spec / version / count / packaging\n- seller/store type\n- shipping or delivery note\n- URL or screenshot context\n\nIf the baseline is only a title:\n- compare competitor platforms against the title identity\n- say `淘宝基准价未提供，本次只比较竞品平台可见结果`\n- avoid saying `值得从淘宝换平台` unless the user later provides Taobao price or visible Taobao evidence\n\nIf the user provides a Taobao price:\n- call it `用户提供的淘宝基准价`\n- compare it to normalized competitor prices\n- flag that final payable price may change with address, coupon eligibility, account status, and stock\n\n## Browser Execution Rules\n\n- Prefer the isolated OpenClaw browser unless the user explicitly asks to use their Chrome tab.\n- Start with one tab per site when practical.\n- Re-snapshot after navigation or major DOM changes.\n- If a site shows login walls, anti-bot interstitials, region prompts, or app-download overlays, use the visible web result if possible and mention the limitation.\n- If a site blocks access completely, report it instead of trying to bypass it.\n- Do not fabricate missing prices.\n- Prefer visible search/listing pages over deep product pages when one platform is unstable.\n- Capture enough evidence to justify the recommendation, not just enough to fill a table.\n\n## Search Targets\n\nUse the standard web search pages when possible:\n\n- 京东: search for the product name on jd.com\n- 拼多多: search for the product name on pinduoduo.com or the visible web listing/search experience available in browser\n- 唯品会: search for the product name on vip.com\n\nIf direct site search is unstable in browser, use a public search engine query constrained to the site, then open the most relevant visible result. Example pattern:\n- `site:jd.com 商品名`\n- `site:pinduoduo.com 商品名`\n- `site:vip.com 商品名`\n\nStill use browser navigation for the actual evidence collection.\n\n## Matching Score\n\nTreat listings as comparable only when the core attributes align.\n\nScore each candidate out of 100:\n- Brand: 25\n- Product line / model / generation: 25\n- Variant and must-match specs: 20\n- Size / weight / count / packaging: 15\n- Seller/channel trust when relevant: 10\n- Evidence completeness: 5\n\nUse these match bands:\n- `高` / 85-100: same item or same SKU-equivalent listing; price comparison can drive the recommendation.\n- `中` / 65-84: near match with one explainable difference; include it with caveats, and make only a weak recommendation unless the tradeoff is obvious.\n- `低` / below 65: reference or substitute only; do not call it cheaper than Taobao or cheaper than another exact match.\n\nForce the match band to `低` when any critical mismatch appears:\n- different model, generation, storage, capacity, size, flavor, color, or pack count\n- refurbished / open-box / parallel import when the baseline is standard new domestic stock\n- trial/sample size compared with regular size\n- unclear warranty or authenticity for safety-sensitive or high-counterfeit-risk categories\n- missing enough information to confirm the listing is the same item\n\nIf there is no close match on a platform:\n- say `未找到足够接近的同款`\n- optionally include the nearest visible alternative, clearly labeled as `近似款` or `替代款`\n\n## Price Normalization Rules\n\nDo not rank prices until they are normalized.\n\nFor each candidate, separate these fields when visible:\n- `标价`: the headline displayed price\n- `可见到手价`: coupon-after, subsidy, member, or promo price, only when the condition is visible\n- `优惠条件`: coupon, membership, group-buy, payment method, limited-time sale, regional subsidy, or account qualification\n- `单位基准`: per item, per pack, per 100g, per ml, per sheet, per pair, or another meaningful unit\n- `运费/门槛`: shipping fee, free-shipping threshold, installation fee, or delivery limitation\n\nUse the most conservative comparable price:\n- Prefer unconditional visible price when coupon eligibility is unclear.\n- Use coupon-after price only when the page makes the condition visible and likely attainable.\n- Keep member-only, group-buy, subsidy, bank-card, and region-dependent prices separate from normal prices.\n- For multi-pack goods, compare unit price and total package value, not only headline price.\n- For electronics, apparel, beauty, and branded goods, do not let a lower price override warranty, channel, authenticity, or version mismatch.\n\nNever recommend a cheaper platform when the cheaper listing is only cheaper because of:\n- lower specification\n- different quantity or packaging\n- unclear seller trust\n- member-only or coupon-after price not available to most users\n- group-buy requirement the user may not want\n- missing shipping, installation, warranty, or invoice cost\n\n## Decision Rules\n\nUse this order of judgment:\n\n1. Is the candidate the same item or only a near match\n2. Is the visible price directly comparable after normalization\n3. Is the seller/store trust level similar\n4. Are coupon, membership, subsidy, or group-buy conditions required\n5. Are shipping speed, warranty, invoice, return rights, and after-sales meaningfully different\n6. Does the user's likely preference favor lowest price, trust, delivery speed, or exact variant availability\n\nRecommendation strength:\n- `强推荐`: high match, clear normalized price, trustworthy seller/channel, and the advantage is material.\n- `弱推荐`: high or medium match, but with one meaningful caveat such as coupon dependency, seller difference, delivery uncertainty, or missing Taobao baseline price.\n- `仅供参考`: partial evidence, noisy results, medium/low match, or blocked platform pages.\n- `无法判断`: no comparable listings, missing baseline for the user question, or evidence too incomplete to support a purchase decision.\n\nIf Taobao is not actually the best option, say so directly. If none of the results are truly comparable, explicitly say `当前不适合做强结论`.\n\n## What To Capture Per Site\n\nCapture only information visible on the page. Prefer the first 1-3 relevant results.\n\nFor each selected listing, collect when visible:\n- platform\n- title\n- displayed price\n- visible final/promo price and condition\n- unit price or normalized price basis, when applicable\n- package/specification\n- store/seller name\n- delivery or shipping note\n- URL\n- match score and band\n- confidence: high / medium / low\n- note about why it matches or why it is only approximate\n\nAlso capture, when visible and relevant:\n- official/self-operated/flagship indicator\n- coupon or subsidy dependency\n- group-buy requirement\n- delivery speed or shipping promise\n- return, warranty, invoice, or authenticity guarantee\n\nEvidence quality gate:\n- A candidate should include at least title, displayed price, spec/version, and URL to drive a recommendation.\n- If two or more of those fields are missing, downgrade confidence to low.\n- Low-confidence candidates can appear in the table but cannot be the sole basis for `强推荐`.\n\n## Output Format\n\nReturn a decision first, then the evidence table.\n\nStart with a short verdict block:\n\n- `淘宝基准`: visible/user-provided price, or `淘宝基准价未提供`\n- `推荐平台`: platform or `暂不建议换平台`\n- `推荐强度`: `强推荐` / `弱推荐` / `仅供参考` / `无法判断`\n- `最低可见可比价`: platform + normalized price basis, only among comparable items\n- `值不值得换平台`: yes / no / depends, with one sentence\n- `主要原因`\n- `风险点`\n- `核验时间`: include date/time or say browser-visible at time of checking\n\nThen return a concise comparison table and short notes.\n\nUse a table like this:\n\n| 平台 | 商品标题 | 标价 | 可见到手价/条件 | 单位价/基准 | 规格/版本 | 店铺 | 匹配分 | 匹配度 | 备注 |\n|---|---|---:|---|---|---|---|---:|---|---|\n| 淘宝 | ... | ¥... | ... | ... | ... | ... | 基准 | 基准 | 用户提供/页面可见 |\n| 京东 | ... | ¥... | ... | ... | ... | ... | 92 | 高 | 同品牌同规格 |\n| 拼多多 | ... | ¥... | ... | ... | ... | ... | 78 | 中 | 规格接近，包装不同 |\n| 唯品会 | ... | ¥... | ... | ... | ... | ... | 52 | 低 | 仅找到近似款 |\n\nThen add:\n- `最低可见价`: platform and raw visible price\n- `最低可比价`: platform and normalized comparable price\n- `可比性判断`: high / medium / low, with reason\n- `风险提示`: differences in package, seller, promo timing, membership price, shipping, warranty, invoice, or coupon requirements\n- `购买建议`: 直接买 / 可等等 / 只建议在某平台买 / 暂不建议下单\n- `下单前核对`: final payable price, address-based delivery, coupon eligibility, stock, warranty, and return policy\n\n## Interpretation Rules\n\n- Do not claim a platform is cheaper unless the compared items are materially comparable.\n- Separate `标价` from coupon-after price when the page makes that distinction.\n- Mention when a price may depend on membership, flash sale, subsidy, payment method, group-buy, or region.\n- If search results are noisy, prefer accuracy over completeness.\n- If the Taobao baseline price is missing, do not say the user should switch away from Taobao; say which competitor has the best visible comparable offer.\n- If the Taobao baseline is not actually the best option, say so directly.\n- If none of the results are truly comparable, explicitly say `当前不适合做强结论`.\n- If the user seems purchase-ready, optimize the answer for actionability: where to buy, what to verify before paying, and what tradeoff they are accepting.\n\n## Example Requests\n\n- `帮我查一下“德芙黑巧克力 84g”在京东、拼多多、唯品会的价格`\n- `对比一下“iPhone 16 Pro 256GB”在几个平台上的可见报价`\n- `把这个淘宝商品名拿去京东、拼多多、唯品会搜同款，做个价格表`\n- `这个淘宝商品有没有更便宜但靠谱的平台`\n- `帮我判断这件商品有没有必要从淘宝换到京东买`\n- `淘宝价 129，这个商品换平台买值不值`\n- `别只比价，也告诉我哪个平台更值得下单`\n\n## Failure Handling\n\nIf one or more sites cannot be accessed or searched reliably, still return a partial result and list:\n- which site failed\n- what was attempted\n- whether the failure was due to login wall, anti-bot page, timeout, app-only flow, or missing web search results\n\nIf evidence is partial, downgrade recommendation strength.\n\nDo not fill missing evidence with guesses. Use:\n- `未见明确价格`\n- `未见规格`\n- `未见店铺信息`\n- `无法确认同款`\n\n## Resource\n\n- Read `references/site-notes.md` when you need execution reminders for JD, Pinduoduo, and Vipshop search behavior, evidence standards, normalized pricing, and recommendation gates.\n\nFile v1.1.0:README.md\n\n# Taobao Competitor Analyzer\n\nCross-platform shopping decision skill for Taobao users.\n\nThis skill checks the same or closest-matching product on:\n- JD.com\n- Pinduoduo\n- Vipshop\n\nThen it answers the question users actually care about:\n\n`Should I keep buying this on Taobao, switch platforms, wait, or avoid this deal?`\n\n## What It Does\n\n- Compares visible prices across major Chinese marketplaces\n- Uses Taobao as the baseline when a Taobao price, link, screenshot, or visible listing is available\n- Matches brand, model, spec, count, and packaging before comparing price\n- Scores same-item confidence so near matches do not masquerade as cheaper exact matches\n- Normalizes headline price, coupon-after price, member price, subsidy price, group-buy price, shipping, and unit price\n- Flags near-match risk instead of pretending similar items are identical\n- Weighs seller trust, shipping, warranty, invoice, and after-sales guarantees\n- Returns a recommendation strength: strong, weak, reference-only, or cannot judge\n\n## v1.1.0 Highlights\n\n- Taobao baseline handling: no baseline price means no unsupported \"switch away from Taobao\" claim\n- 100-point match scoring across brand, model, specs, quantity, channel trust, and evidence completeness\n- Price normalization for unit price, coupon dependency, membership, subsidy, group-buy, and shipping caveats\n- Evidence quality gate for title, price, spec/version, and URL before a listing can drive a strong recommendation\n- Expanded verdict block with recommendation strength, lowest comparable price, risks, and final checks before purchase\n\n## Best Use Cases\n\n- Compare a Taobao listing with JD / PDD / Vipshop\n- Check whether a \"cheaper\" result is actually the same SKU\n- Decide if it is worth switching platforms\n- Do fast competitor research for consumer products\n- Find the lowest visible comparable price with caveats\n- Separate normal price from coupon, member, subsidy, and group-buy conditions\n\n## Example Prompts\n\n- `帮我查这个淘宝商品在京东、拼多多、唯品会有没有同款`\n- `淘宝价 129，这个商品换到京东买值不值`\n- `别只比价，告诉我哪个平台更值得买`\n- `Compare this Taobao product with JD, PDD, and Vipshop and recommend where to buy`\n\n## Output Style\n\nThe skill is optimized to return:\n- Taobao baseline status\n- recommended platform\n- recommendation strength\n- lowest visible comparable price\n- normalized unit/condition basis\n- whether the comparison is apples-to-apples\n- major risks and caveats\n- a direct buy / wait / avoid recommendation\n\n## Positioning\n\nThis is not a generic shopping browser helper.\n\nIt is a focused purchase-decision skill for users who want:\n- faster price comparison\n- fewer fake \"cheap\" matches\n- clearer marketplace tradeoff analysis\n- conservative recommendations when evidence is incomplete\n\nFile v1.1.0:_meta.json\n\n{\n  \"ownerId\": \"kn77zzg9p845zanvy6vrf76k7d81mcnm\",\n  \"slug\": \"taobao-competitor-analyzer\",\n  \"version\": \"1.1.0\",\n  \"publishedAt\": 1780836553691\n}\n\nFile v1.1.0:references/site-notes.md\n\n# Site Notes\n\n## Goal\n\nUse browser-visible marketplace pages to collect price evidence for the same or closest-matching product on JD, Pinduoduo, and Vipshop, then turn that evidence into a purchase decision against the Taobao baseline when available.\n\n## Common Reminders\n\n- Prefer visible desktop web pages.\n- Use exact product names first, then a lightly simplified query if results are sparse.\n- Compare like-for-like products only.\n- Keep the evidence chain simple: search page -> listing -> visible title/price/spec.\n- Do not use APIs, hidden JSON endpoints, or scraping shortcuts outside the browser tool.\n- Price is not enough; capture trust, conditions, and caveats.\n- The final answer should help the user choose, not just browse.\n\n## Baseline Discipline\n\n- Treat Taobao as the baseline only when a Taobao price, URL, screenshot, or user-provided listing detail exists.\n- If the user provides only a Taobao title, search against that identity but write `淘宝基准价未提供`.\n- Do not claim the user should switch away from Taobao unless Taobao's visible or user-provided price is known.\n- When the user provides the Taobao price, label it as user-provided unless verified in browser.\n\n## Matching Score Reminders\n\nUse the 100-point score from `SKILL.md`:\n- brand: 25\n- model / line / generation: 25\n- variant and must-match specs: 20\n- size / weight / count / packaging: 15\n- seller or channel trust: 10\n- evidence completeness: 5\n\nInterpretation:\n- 85-100: high, can drive recommendation\n- 65-84: medium, caveated recommendation only\n- below 65: low, reference/near-match only\n\nDowngrade to low when the result has a critical mismatch such as different generation, capacity, count, refurbished status, unclear warranty, trial size, or unverified authenticity in a risk-sensitive category.\n\n## Price Normalization Reminders\n\nBefore ranking, separate:\n- headline/list price\n- visible final or coupon-after price\n- coupon/member/subsidy/group-buy/payment/region conditions\n- unit basis such as per pack, per sheet, per 100g, per ml, per pair, or per item\n- shipping, free-shipping threshold, installation, warranty, invoice, or delivery constraints\n\nUse conservative comparison:\n- Prefer unconditional visible price when promo eligibility is unclear.\n- Treat member-only, group-buy, payment, and regional subsidy prices as conditional.\n- For multi-pack daily goods, compare unit price and total quantity.\n- For electronics, beauty, apparel, infant goods, medical/safety goods, and branded products, keep authenticity, warranty, and channel trust ahead of headline price.\n\n## JD\n\n- Usually supports standard web search and listing pages well.\n- Prefer self-operated or official flagship listings when multiple near-identical results exist.\n- Watch for coupon text, plus/member pricing, trade-in, bank/payment offers, and promotional banners.\n- Distinguish list price from final promo price if both are shown.\n- Treat 京东自营 and official flagship stores as stronger trust signals when prices are close.\n- For electronics, appliances, beauty, infant goods, and regulated products, do not recommend a weaker channel solely because it is cheaper.\n\n## Pinduoduo\n\n- Web results may be noisier than JD.\n- Watch for subsidy labels, group-buy wording, coupon claims, and strong promo framing.\n- Matching confidence should be reduced when the seller, spec, or packaging is unclear.\n- If the browser experience is limited, use a public search engine with site restriction, then open the visible result page.\n- If the low price depends on 拼团 or 百亿补贴, call that out explicitly in the final recommendation.\n- Reduce recommendation strength when seller trust is unclear, even if the price looks excellent.\n- A very low PDD price should be treated as a lead to verify, not as a conclusion, unless match and seller evidence are strong.\n\n## Vipshop\n\n- Results may lean toward branded discount inventory and variant-specific listings.\n- Pay attention to size/color/version because discount channels often surface adjacent variants.\n- If the exact match is missing, mark the result as `近似款` instead of treating it as the same SKU.\n- Vipshop often wins on branded discounts but loses on exact spec coverage; do not overstate comparability.\n- Treat limited stock, size gaps, color mismatch, and flash-sale timing as decision caveats.\n\n## Evidence Standard\n\nFor each platform, aim to capture:\n- title\n- displayed price\n- visible final price or promo condition\n- variant/specification\n- seller/store if visible\n- URL\n- matching score and reason\n- a short matching note\n\nWhen visible, also capture:\n- official / self-operated / flagship status\n- coupon or membership dependency\n- subsidy / group-buy dependency\n- shipping promise\n- return / warranty / invoice / authenticity guarantee\n\nEvidence quality gate:\n- A listing should have title, price, spec/version, and URL before it can support a strong recommendation.\n- If two or more of those are missing, downgrade confidence to low.\n- Low-confidence candidates can be included, but they should not determine the final answer alone.\n\n## Output Discipline\n\nUse short, decision-friendly language.\n\nRecommended summary fields:\n- Taobao baseline status\n- lowest visible raw price\n- lowest visible comparable price\n- whether the comparison is apples-to-apples\n- major caveats: variant mismatch, coupon dependency, seller difference, shipping difference, warranty/invoice difference, regional limitation\n- recommended platform\n- recommendation strength: strong / weak / reference only / cannot judge\n- one-sentence reason the user should or should not switch away from Taobao\n- final checks before buying: final payable price, address-based delivery, coupon eligibility, stock, warranty, and return policy\n\nFile v1.1.0:skill-card.md\n\n## Description: <br>\nCompare a Taobao product with JD.com, Pinduoduo, and Vipshop using browser-visible evidence only, normalize prices, score same-item confidence, and tell the user where it is actually worth buying. <br>\n\nThis skill is ready for commercial/non-commercial use. <br>\n\n## Publisher: <br>\n[harrylabsj](https://clawhub.ai/user/harrylabsj) <br>\n\n### License/Terms of Use: <br>\nMIT-0 <br>\n\n\n## Use Case: <br>\nExternal shopping researchers and Taobao buyers use this skill to compare a Taobao product against visible JD.com, Pinduoduo, and Vipshop results, normalize prices and conditions, and decide whether to buy, wait, avoid, or stay with Taobao. <br>\n\n### Deployment Geography for Use: <br>\nGlobal <br>\n\n## Known Risks and Mitigations: <br>\nRisk: Account-specific prices, coupons, delivery options, or shopping history may be exposed when using a personal logged-in browser. <br>\nMitigation: Use the isolated browser for normal comparisons; use a personal logged-in browser only when the user accepts account-context exposure. <br>\nRisk: Visible marketplace prices can depend on coupons, memberships, subsidies, group-buy terms, region, shipping, stock, or account eligibility. <br>\nMitigation: Separate headline price from conditional final price and ask the user to verify final payable price, delivery address, coupon eligibility, warranty, stock, and return terms before buying. <br>\nRisk: Near-match listings may appear cheaper while differing by model, specification, package count, seller trust, warranty, or authenticity signals. <br>\nMitigation: Use same-item match scoring and downgrade recommendation strength when critical attributes or evidence are incomplete. <br>\n\n\n## Reference(s): <br>\n- [ClawHub listing](https://clawhub.ai/harrylabsj/taobao-competitor-analyzer) <br>\n- [Mark\n\nArchive v1.0.3: 4 files, 7415 bytes\n\nFiles: README.md (1760b), references/site-notes.md (2867b), SKILL.md (8973b), _meta.json (145b)\n\nArchive v1.0.2: 4 files, 7200 bytes\n\nFiles: README.md (1760b), references/site-notes.md (2867b), SKILL.md (8454b), _meta.json (145b)\n\nArchive v1.0.1: 3 files, 6013 bytes\n\nFiles: references/site-notes.md (2867b), SKILL.md (8275b), _meta.json (145b)\n\nArchive v1.0.0: 3 files, 4184 bytes\n\nFiles: references/site-notes.md (1910b), SKILL.md (5196b), _meta.json (145b)","readmeExcerpt":"Skill: Taobao Price Compare Owner: harrylabsj Summary: Taobao price and competitor comparison assistant. Input a Taobao title/link or product need; compare JD, PDD, and Vipshop using browser-visible evidence, nor... Tags: buying-advice:2.1.1, category-aware:2.1.1, china:2.1.1, comparison:2.1.1, competitor-analysis:2.1.1, ecommerce:2.1.1, latest:2.1.1, price-analysis:1.2.0, price-compare:2.1.1, shopping:2.1.1, shoppin","codeSnippets":[],"executableExamples":[],"parameters":null,"dependencies":[],"permissions":[],"extractedFiles":[{"path":"SKILL.md","content":"---\nname: taobao-competitor-analyzer\ndescription: \"Taobao price and competitor comparison assistant. Input a Taobao title/link or product need; compare JD, PDD, and Vipshop using browser-visible evidence, normalize prices, score same-item confidence, and output where to buy. Safe boundary: no login, no order submission, no payment.\"\n---\n# Taobao Competitor Analyzer\n\nCompare a Taobao product with the same or closest-matching listings on 京东、拼多多、唯品会 using the browser tool only. Work from a product name, Taobao title, link, screenshot, or user-provided baseline details. Return a compact purchase decision with visible price evidence, normalized comparability, match score, category-aware platform fit, recommendation strength, and risk notes.\n\nWhat makes this skill useful:\n- It compares comparable items instead of chasing misleading low prices.\n- It separates Taobao baseline price from competitor prices.\n- It normalizes price by spec, quantity, shipping, and promo conditions before ranking.\n- It scores same-item confidence before making a buying recommendation.\n- It applies category-specific risk thresholds instead of treating shoes, phones, skincare, and paper towels the same way.\n- It explains whether a lower price depends on coupons, membership, subsidy, or group-buy.\n- It ends with a clear buy / wait / avoid recommendation instead of just a table.\n\n## When To Use\n\nUse this skill when the user is effectively asking:\n- 这件淘宝商品别的平台多少钱\n- 有没有同款或更划算的平台\n- 京东 / 拼多多 / 唯品会 哪个更值得买\n- 这几个平台价格差这么多正常吗\n- 帮我做一个同款比价和购买建议\n- 这个淘宝商品换平台买值不值\n- 同样价差下哪个平台风险更低\n\nThe skill should optimize for purchase decisions, not raw data collection.\n\n## Commerce Matrix\n\nThis skill is the cross-platform comparison node in the shopping matrix.\n\nPrefer nearby skills when the task is narrower:\n- `taobao-shopping` for Taobao-only listing and seller evaluation\n- `jd-shopping` for trust-first self-operated buying\n- `pdd-shopping` for low-price and subsidy-first buying\n- `tianmao` for flagship-store and authenticity-first buying\n- `vip` for branded discount and flash-sale buying\n- `alibaba-shopping` when the user first needs to choose between Taobao, Tmall, and 1688\n\n## Taobao Baseline Bridge\n\nUse `taobao-shopping` first, or apply its same discipline yourself, when the Taobao side is not yet clear.\n\nBefore cross-platform comparison, establish the Taobao baseline:\n\n- exact product identity\n- selected SKU attributes\n- visible or user-provided Taobao price\n- store type and seller trust cue\n- coupon / membership / threshold condition\n- return, warranty, invoice, authenticity, or delivery caveat when relevant\n\nIf the user only provides a Taobao title with no price or listing evidence, this skill can compare competitor visible offers, but it must not claim the user should switch away from Taobao. Write `淘宝基准价未提供`.\n\nIf the user's real question is only \"is this Taobao listing trustworthy?\", stay in or route to `taobao-shopping` instead of broadening into JD/PDD/Vipshop.\n\n## Workflow\n\n1. Normalize the Taoba"},{"path":"README.md","content":"# Taobao Competitor Analyzer\n\nCross-platform shopping decision skill for Taobao users.\n\nThis skill checks the same or closest-matching product on:\n- JD.com\n- Pinduoduo\n- Vipshop\n\nThen it answers the question users actually care about:\n\n`Should I keep buying this on Taobao, switch platforms, wait, or avoid this deal?`\n\n## What It Does\n\n- Compares visible prices across major Chinese marketplaces\n- Uses Taobao as the baseline when a Taobao price, link, screenshot, or visible listing is available\n- Matches brand, model, spec, count, and packaging before comparing price\n- Scores same-item confidence so near matches do not masquerade as cheaper exact matches\n- Normalizes headline price, coupon-after price, member price, subsidy price, group-buy price, shipping, and unit price\n- Applies category playbooks so phones, skincare, baby goods, shoes, snacks, and paper towels are judged differently\n- Flags near-match risk instead of pretending similar items are identical\n- Weighs seller trust, shipping, warranty, invoice, and after-sales guarantees\n- Returns a recommendation strength: strong, weak, reference-only, or cannot judge\n\n## v1.1.0 Highlights\n\n- Taobao baseline handling: no baseline price means no unsupported \"switch away from Taobao\" claim\n- 100-point match scoring across brand, model, specs, quantity, channel trust, and evidence completeness\n- Price normalization for unit price, coupon dependency, membership, subsidy, group-buy, and shipping caveats\n- Evidence quality gate for title, price, spec/version, and URL before a listing can drive a strong recommendation\n- Expanded verdict block with recommendation strength, lowest comparable price, risks, and final checks before purchase\n\n## v1.2.0 Highlights\n\n- Category-aware playbooks for electronics/appliances, beauty/personal care, food/daily goods, apparel/shoes/bags, baby/health/safety, and low-risk standard goods\n- Buying-posture routing for lowest price, authenticity/after-sales, delivery speed, exact variant, and research-first workflows\n- Platform-fit judgment so JD, PDD, and Vipshop are weighed differently by category instead of by headline price alone\n- Category-specific price-delta thresholds: low-risk commodities can switch on unit-price wins, while warranty/authenticity categories need stronger evidence\n- New `品类策略` and `购买偏好` fields in the verdict block\n\n## Best Use Cases\n\n- Compare a Taobao listing with JD / PDD / Vipshop\n- Check whether a \"cheaper\" result is actually the same SKU\n- Decide if it is worth switching platforms\n- Do fast competitor research for consumer products\n- Find the lowest visible comparable price with caveats\n- Separate normal price from coupon, member, subsidy, and group-buy conditions\n- Decide whether a low price is worth the category-specific risk\n\n## Example Prompts\n\n- `帮我查这个淘宝商品在京东、拼多多、唯品会有没有同款`\n- `淘宝价 129，这个商品换到京东买值不值`\n- `这款护肤品拼多多便宜很多，风险大不大`\n- `这双鞋唯品会有近似款，能不能买`\n- `别只比价，告诉我哪个平台更值得买`\n- `Compare this Taobao product with JD, PDD, and Vipshop and recommend where to buy"},{"path":"_meta.json","content":"{\n  \"ownerId\": \"kn77zzg9p845zanvy6vrf76k7d81mcnm\",\n  \"slug\": \"taobao-competitor-analyzer\",\n  \"version\": \"2.1.1\",\n  \"publishedAt\": 1782214405857\n}"},{"path":"references/category-playbooks.md","content":"# Category Playbooks\n\n## Goal\n\nUse these playbooks after normalizing the Taobao baseline and before making the final recommendation. The goal is to avoid one-size-fits-all price sorting. A lower price means different things for phones, skincare, baby goods, shoes, snacks, and paper towels.\n\n## How To Use\n\n1. Classify the product into the closest category.\n2. Infer the user's buying posture: low price, trust/after-sales, delivery speed, exact variant, or research.\n3. Apply the category's must-check attributes and risk gates.\n4. Decide whether the price gap is large enough to overcome category risk.\n5. State the playbook in the verdict as `品类策略`.\n\nIf the category is unclear, write the assumption and keep recommendation strength at `弱推荐` or lower.\n\n## Buying Postures\n\n### Low Price First\n\n- Use normalized unit price after matching gates pass.\n- Prefer unconditional visible prices over conditional coupon/member/group-buy prices.\n- Strong recommendations still need high match and enough seller/channel evidence.\n\n### Trust And After-Sales First\n\n- Prefer official, self-operated, flagship, or authorized channels.\n- Treat invoice, warranty, returns, and authenticity evidence as decision drivers.\n- A cheaper weak-channel listing usually becomes `弱推荐` or `仅供参考`.\n\n### Speed First\n\n- Prefer visible delivery promise, local stock, self-operated logistics, and stable fulfillment.\n- Downgrade group-buy, pre-sale, limited stock, and unclear shipping.\n- If delivery is address-dependent and not visible, include it in `下单前核对`.\n\n### Exact Variant First\n\n- Prioritize exact model, color, size, flavor, version, pack count, batch, and bundle contents.\n- Do not treat near matches as cheaper exact matches.\n- For apparel, shoes, beauty, and electronics, variant mismatch usually blocks `强推荐`.\n\n### Research First\n\n- Include exact matches, near matches, and substitutes in separate groups.\n- Do not rank substitutes against exact matches as if they are identical.\n- Use this mode for marketplace research, category scans, and competitor mapping.\n\n## Electronics And Appliances\n\nExamples: phones, headphones, laptops, cameras, routers, appliances, smart devices.\n\nMust check:\n- exact model, generation, storage/capacity, color, region/version, bundle, and warranty\n- new vs refurbished/open-box/parallel import\n- official/self-operated/authorized channel\n- invoice, warranty, installation, return policy, trade-in, and delivery timing\n\nPlatform fit:\n- JD is often stronger when self-operated, official, invoice, warranty, fast delivery, or installation matters.\n- PDD can be considered when subsidy/official evidence is visible and the SKU is exact.\n- Vipshop is usually weaker for exact electronics coverage unless the listing is clearly official and exact.\n\nRecommendation gate:\n- Do not recommend a weaker channel for small savings.\n- Require high match and clear warranty/channel evidence for `强推荐`.\n- If savings are modest and JD/Taobao official support is stronger, recommend staying with the s"},{"path":"references/site-notes.md","content":"# Site Notes\n\n## Goal\n\nUse browser-visible marketplace pages to collect price evidence for the same or closest-matching product on JD, Pinduoduo, and Vipshop, then turn that evidence into a purchase decision against the Taobao baseline when available.\n\n## Common Reminders\n\n- Prefer visible desktop web pages.\n- Use exact product names first, then a lightly simplified query if results are sparse.\n- Compare like-for-like products only.\n- Keep the evidence chain simple: search page -> listing -> visible title/price/spec.\n- Do not use APIs, hidden JSON endpoints, or scraping shortcuts outside the browser tool.\n- Price is not enough; capture trust, conditions, and caveats.\n- The final answer should help the user choose, not just browse.\n- Read `category-playbooks.md` when the product category changes how much price should matter.\n\n## Baseline Discipline\n\n- Treat Taobao as the baseline only when a Taobao price, URL, screenshot, or user-provided listing detail exists.\n- If the user provides only a Taobao title, search against that identity but write `淘宝基准价未提供`.\n- Do not claim the user should switch away from Taobao unless Taobao's visible or user-provided price is known.\n- When the user provides the Taobao price, label it as user-provided unless verified in browser.\n\n## Matching Score Reminders\n\nUse the 100-point score from `SKILL.md`:\n- brand: 25\n- model / line / generation: 25\n- variant and must-match specs: 20\n- size / weight / count / packaging: 15\n- seller or channel trust: 10\n- evidence completeness: 5\n\nInterpretation:\n- 85-100: high, can drive recommendation\n- 65-84: medium, caveated recommendation only\n- below 65: low, reference/near-match only\n\nDowngrade to low when the result has a critical mismatch such as different generation, capacity, count, refurbished status, unclear warranty, trial size, or unverified authenticity in a risk-sensitive category.\n\n## Price Normalization Reminders\n\nBefore ranking, separate:\n- headline/list price\n- visible final or coupon-after price\n- coupon/member/subsidy/group-buy/payment/region conditions\n- unit basis such as per pack, per sheet, per 100g, per ml, per pair, or per item\n- shipping, free-shipping threshold, installation, warranty, invoice, or delivery constraints\n\nUse conservative comparison:\n- Prefer unconditional visible price when promo eligibility is unclear.\n- Treat member-only, group-buy, payment, and regional subsidy prices as conditional.\n- For multi-pack daily goods, compare unit price and total quantity.\n- For electronics, beauty, apparel, infant goods, medical/safety goods, and branded products, keep authenticity, warranty, and channel trust ahead of headline price.\n\n## Category Shortcut\n\nUse `references/category-playbooks.md` to decide which risk dominates:\n- warranty and invoice for electronics/appliances\n- authenticity, batch, and expiry for beauty/personal care\n- unit price, pack count, and freshness for food/daily goods\n- size, color, style code, and return policy for apparel/shoes/bags\n- of"}],"languages":[],"docsSourceLabel":"CLAWHUB","editorialOverview":null,"editorialQuality":{"score":100,"threshold":65,"status":"thin","wordCount":2162,"uniquenessScore":37,"reasons":["uniqueness-below-45"]}},"media":{"evidence":{"source":"no-media","verified":false,"confidence":"low","updatedAt":"2026-10-09T16:47:14.132Z","emptyReason":"No screenshots, media assets, or demo links are available."},"primaryImageUrl":null,"mediaAssetCount":0,"assets":[],"demoUrl":null},"ownerResources":{"evidence":{"source":"unclaimed","verified":false,"confidence":"low","updatedAt":"2026-10-09T16:47:14.132Z","emptyReason":"This page has not been claimed by the agent owner."},"hasCustomPage":false,"customPageUpdatedAt":null,"customLinks":[],"structuredLinks":{"docsUrl":null,"demoUrl":null,"supportUrl":null,"pricingUrl":null,"statusUrl":null},"customPage":null},"relatedAgents":{"evidence":{"source":"protocol-neighbors","verified":false,"confidence":"medium","updatedAt":"2026-10-09T18:36:30.859Z","emptyReason":null},"items":[{"id":"b917f68a-ebff-438e-84f8-3f4b2494c0bc","entityType":"agent","canonicalPath":"/agent/activepieces-activepieces","slug":"activepieces-activepieces","name":"activepieces","description":"AI Agents & MCPs & AI Workflow Automation • (~400 MCP servers for AI agents) • AI Automation / AI Agent with MCPs • AI Workflows & AI Agents • MCPs for AI Agents","url":"https://github.com/activepieces/activepieces","homepage":"https://www.activepieces.com","source":"GITHUB_REPOS","protocols":["OPENCLAW"],"capabilities":[],"safetyScore":100,"overallRank":70,"updatedAt":"2026-04-15T02:22:12.426Z","createdAt":"2026-02-25T03:38:12.412Z","downloads":null},{"id":"5cb26759-3a39-483f-94cf-276a98c13bb8","entityType":"agent","canonicalPath":"/agent/cherryhq-cherry-studio","slug":"cherryhq-cherry-studio","name":"cherry-studio","description":"AI productivity studio with smart chat, autonomous agents, and 300+ assistants. Unified access to frontier LLMs","url":"https://github.com/CherryHQ/cherry-studio","homepage":"https://cherry-ai.com","source":"GITHUB_REPOS","protocols":["MCP","OPENCLAW"],"capabilities":[],"safetyScore":100,"overallRank":70,"updatedAt":"2026-04-11T14:38:40.986Z","createdAt":"2026-02-25T03:38:19.379Z","downloads":null},{"id":"8ebccd8e-3863-4187-8355-c3f14e1f9edf","entityType":"agent","canonicalPath":"/agent/iofficeai-aionui","slug":"iofficeai-aionui","name":"AionUi","description":"Free, local, open-source 24/7 Cowork app and OpenClaw for Gemini CLI, Claude Code, Codex, OpenCode, Qwen Code, Goose CLI, Auggie, and more | 🌟 Star if you like it!","url":"https://github.com/iOfficeAI/AionUi","homepage":"https://www.aionui.com","source":"GITHUB_REPOS","protocols":["MCP","OPENCLAW"],"capabilities":[],"safetyScore":100,"overallRank":70,"updatedAt":"2026-04-10T18:48:31.762Z","createdAt":"2026-02-25T03:38:16.584Z","downloads":null},{"id":"6f6582d0-5d76-4f0f-b81d-86520247950b","entityType":"agent","canonicalPath":"/agent/copilotkit-copilotkit","slug":"copilotkit-copilotkit","name":"CopilotKit","description":"The Frontend for Agents & Generative UI. React + Angular","url":"https://github.com/CopilotKit/CopilotKit","homepage":"https://docs.copilotkit.ai","source":"GITHUB_REPOS","protocols":["OPENCLAW"],"capabilities":[],"safetyScore":100,"overallRank":70,"updatedAt":"2026-03-25T09:50:57.846Z","createdAt":"2026-02-25T03:39:14.617Z","downloads":null}],"links":{"hub":"/agent","source":"/agent/source/clawhub","protocols":[{"label":"OpenClaw","href":"/agent/protocol/openclew"}]}}}