flipintel

Use these tools to answer business questions about ebay sales performance, product rankings, and fee impact.

От сообществаРаботаетБез входаГлобальныйБесплатноМожет изменять данные

Что умеет

Sales Summary: Revenue, order count, units sold, and average order value for a period. period options: today, yesterday, last_7_days, last_30_days, last_90_days, this_month
Top Products: Top performing listings ranked by revenue, units, or orders. metric options: revenue, units, orders
Revenue Trend: Month-over-month revenue for the last N months with % change. Shows whether your store is growing or declining. months: 2–12

Какие данные видит

Нужен ли аккаунт

Не нужен: сервер работает без входа

Use these tools to answer business questions about ebay sales performance, product rankings, and fee impact. You can even create bundles based on active listings! Always present numbers with context — compare to prior periods when possible and surface the 'so what'.

FlipIntel the tool for ebay sellers!

Список инструментов сервера (17)

Технические названия из tools/list. Нужны только разработчикам.

sales_summary	Revenue, order count, units sold, and average order value for a period. period options: today, yesterday, last_7_days, last_30_days, last_90_days, this_month
top_products	Top performing listings ranked by revenue, units, or orders. metric options: revenue, units, orders
revenue_trend	Month-over-month revenue for the last N months with % change. Shows whether your store is growing or declining. months: 2–12
list_my_listings	Show the seller's currently-active eBay listings — title, price, listing ID, and days active. Read-only. Includes listings with ZERO recent traffic (new items, slow movers, items not yet indexed by eBay search). Use this when the seller asks any variation of: • "What's currently listed in my store?" • "Show me my active listings." • "Did my new listing go live?" • "Which items have I been listing the longest?" This is the only tool that surfaces brand-new listings — every other inventory-related tool (slow_moving_inventory, top_products, conversion_ report, suggest_bundles, get_active_listings) is derived from eBay's analytics traffic_report, which excludes items with zero impressions. ⚠ PAGINATED. Default 10 listings per call (hard cap 25). For stores with many listings, paginate via the response's `has_more` and `next_offset` fields. Fields per listing: listing_id, title, price (formatted with currency), days_active (whole days since the listing went live). Traffic stats (impressions/views/sales) are intentionally NOT included here — see slow_moving_inventory, conversion_report, or top_products for those. Available on all paid tiers.
slow_moving_inventory	Listings with views but zero sales — stuck inventory that needs action. Answers: 'What am I listing that nobody is buying?' ⚠ PAGINATED. Returns at most 10 listings per call (max 25) to keep memory usage bounded. For sellers with many stuck items, paginate by reading `has_more` and `next_offset` from the response and calling again with `offset=next_offset`. min_views: minimum views threshold to qualify (default 1) limit: items per page (default 10, max 25) offset: skip the first N items in the source list (default 0) Response fields driving pagination: • total: total candidates available across all pages • returned: count in this batch (after liveness filter) • has_more / next_offset: pagination cursor For Growth+ users, each listing row includes title, current price, a clickable photo link, and the listing URL — so Claude can name specific items in the response. For Starter users, only listing IDs are returned (call get_listing_details to enrich, which surfaces the upgrade prompt).
fee_breakdown	Net revenue after all eBay fees, broken down by fee type. Shows effective fee rate so you know what you're actually keeping.
estimated_fees	Estimate eBay fees before listing. Shows final value fee, payment processing, and net amount you'd receive for a given sale price.
fee_anomalies	Orders where eBay fees are unusually high compared to your average rate. Answers: 'Where is eBay taking more than their fair share?' period options: today, yesterday, last_7_days, last_30_days, last_90_days, this_month
conversion_report	Impression → view → sale funnel for all active listings. Identifies high-traffic listings that aren't converting so you know where to focus. months_ago: 0 = current month, 1 = last month, etc. For Growth+ users, top/worst converters carry title/price/photo-link/ listing-url and the insight names the worst converter by title — so the action is "rewrite the title on listing X" instead of "look up item ID Y."
category_breakdown	Revenue, units, and items sold broken down by eBay top-level category. Answers: 'What categories are driving my revenue?' and 'What should I source more of?' period options: today, yesterday, last_7_days, last_30_days, last_90_days, this_month
suggest_ebay_category	Ask eBay's Taxonomy API which leaf category best fits a listing title. Returns the top 5 suggestions with their numeric category_id, the leaf name, and the full breadcrumb path. ⚠ ALWAYS CALL THIS BEFORE create_bundle_listing. create_bundle_listing accepts a category_id and respects it verbatim — if you pass the wrong ID, the listing publishes successfully but lands in a random eBay category (zero conversion). Don't guess category IDs from memory; they change and eBay has thousands. Use this tool to ground the choice in eBay's actual taxonomy. Workflow: 1. Compose the bundle's title (e.g., "Vintage Trinket Box Lot of 9") 2. Call this tool with that title 3. Pick the most appropriate suggestion (usually the first one unless the seller has a strong opinion) 4. Pass that `category_id` to create_bundle_listing Returns: { "query": echo of the title you passed, "suggestions": [ { "category_id": "13837", "category_name": "Boxes, Jars & Tins", "category_path": "Collectibles > Decorative Collectibles > Boxes, Jars & Tins", "confidence": int \| null }, ... ] } Empty suggestions list = eBay couldn't match the title; ask the seller to clarify what the bundle is, or pick a category yourself from their existing store. Pro plan or higher.
get_active_listings	Return a PAGE of currently-active eBay listings for the seller's store with item_id, title, price, category, image URL, and recent traffic stats. Sold/ended/inactive listings are filtered out via Browse API. ⚠ PAGINATED. The tool intentionally returns small batches (default 10, max 25 per call) to keep memory usage bounded on the worker. Sellers with many listings → call repeatedly with increasing `offset`. Workflow for "show me all my listings": 1. Call get_active_listings() with no args → returns first 10 2. Check the response's `has_more` and `next_offset` 3. If `has_more` is true, call get_active_listings(offset=next_offset) 4. Repeat until `has_more` is false Response fields (use them to drive pagination): • total_with_traffic: total candidates available across all pages • offset: starting index of this page • returned: count actually returned in this batch (after liveness filter) • has_more: true if there are more listings to fetch • next_offset: pass this back as `offset` to get the next batch • listings: the actual rows This is the entry point for the bundle creation workflow. Use the returned item_ids with: • generate_bundle_collage — make a composite photo from chosen items • create_bundle_listing — publish a new bundle listing on eBay • end_listings — remove the individual items once the bundle is live Images are returned as URLs only (not embedded). For per-item visual review, chain into get_listing_details with the IDs you want to inspect. limit: items per page (default 10, max 25). Hard-capped — over-asks are clamped to 25. offset: skip the first N items in the source list (default 0). Listings are ordered by recent impressions descending; the underlying traffic_report is cached so paginated calls only re-do enrichment. Pro plan or higher.
generate_bundle_collage	Build a composite bundle preview from 2–9 eBay item IDs, upload it (plus the individual source photos) to Cloudflare R2, and return public HTTPS URLs ready to drop into create_bundle_listing. Use this after the seller has picked the listings they want to bundle. Steps in the Pro bundle workflow: 1. get_active_listings — find candidates 2. suggest_bundles — narrow them into themed groupings 3. (this tool) generate_bundle_collage — assemble + host images 4. create_bundle_listing — publish the bundle on eBay Why R2 hosting (not eBay CDN URLs directly): eBay silently refuses its own CDN URLs in new listings — publishes succeed but render BLANK. R2 gives us external HTTPS URLs that work and cost $0 at our scale. Objects auto-expire after 7 days via bucket lifecycle rule. Returns (collage-only mode): collage_url: public R2 URL for the assembled bundle photo image_urls: [collage_url] — a one-element list. Pass it verbatim to create_bundle_listing's image_urls param. The collage becomes the eBay gallery thumbnail. Sellers wanting per-item supplementary photos can add them manually via eBay's listing editor after publish. item_ids: IDs actually included (in stable sort order) missing_ids: IDs we couldn't fetch (ended/inactive listings) image_count: how many sources made it into the collage width/height: collage dimensions in px (meets eBay's ≥500px floor) spec_version: CollageSpec version used to render the collage expires_at: ISO timestamp; R2 objects deleted after this next_steps: one-line handoff hint for create_bundle_listing item_ids: 2–9 active eBay item IDs. Ended/inactive items are skipped and reported in missing_ids. Order doesn't matter — sorted internally so same set always yields the same collage. Pro plan or higher.
create_bundle_listing	Push a new bundle listing to eBay via the Inventory API. Final step in the Pro-tier bundle workflow: 1. get_active_listings — pick candidates 2. suggest_bundles — narrow into themed groupings 3. generate_bundle_collage — assemble photo (seller hosts manually) 4. (this tool) create_bundle_listing Safe by default: creates an UNPUBLISHED offer (draft) so a typo can't cost real money. The seller reviews on eBay → Inventory and publishes manually. Pass publish=True only when the seller has confirmed.
edit_listing	Revise an already-live eBay listing's title, description, or price. ⚠ TWO-CALL PATTERN. Unlike bundle creation which defaults to an unpublished draft, an edit to a live listing has no buffer — the change is real the moment eBay accepts it. So: Step 1 (default, confirm=False): Call with the proposed changes. The tool fetches the current listing, validates, and returns a before/after diff WITHOUT writing to eBay. Use this to show the seller exactly what will change and get explicit "yes, do it" confirmation in conversation. Step 2 (confirm=True): After the seller has confirmed, call AGAIN with the same args plus confirm=True. The tool re-validates (price could have changed in between), submits the revision via eBay's Trading API ReviseItem, and writes an audit-log row per field changed. Partial update: only fields you pass are changed. Skip a field entirely (leave it None) to leave it untouched.
suggest_bundles	Gather active listings grouped by category so you can propose specific bundle ideas to the seller. Read-only — you suggest, the seller decides what to create on eBay. ⚠ PAGINATED. Returns at most 10 candidate items per call (max 25). For stores with many slow movers, paginate by reading `has_more` / `next_offset` from the first text block and calling again with `offset=next_offset`. Each page produces a coherent set of bundle suggestions; don't try to merge pages into one mega-prompt. The response is text-only: each listing carries title, price, traffic stats, listing URL, and a clickable Photo link the user can open in eBay. Photos are NOT embedded here. Once the seller picks the 2–4 items they want to bundle, chain into `get_listing_details` for embedded thumbnails. Use the data to propose 3–5 specific bundles per page: • A theme/title for the bundle (e.g., "Vintage Glassware Trio") • Which listings go into it (pick 2–4 items from the same category) • Suggested bundle price — usually the sum of individual prices with a 10–20% discount to incentivize the combined purchase • A one-line rationale explaining why these items pair well focus options: • "slow_movers" (default) — listings with views but no recent sales. Best bundle candidates because pairing revives buyer interest. • "all_active" — any listing with recent impressions; use when the seller wants ideas across their whole store. limit: candidates per page (default 10, max 25). Hard-capped. offset: skip the first N candidates in the source list. Growth plan or higher.
get_listing_details	Pull full details for a list of eBay item IDs and return them as a mixed-content response: a text block per listing (title, price, category, condition, watcher count, listing URL) immediately followed by the listing's photo embedded as an inline image. ⚠ INPUT CAPPED at 10 item_ids per call. Each item fetches metadata + a downsampled thumbnail, and the per-call memory cost scales linearly. For a longer list, split it into batches of ≤10 and call repeatedly — the underlying Browse-API responses are cached server-side so repeat calls for the same IDs are cheap. Use this after slow_moving_inventory, top_products, conversion_report, or suggest_bundles when the seller asks "what is item X?" or "show me these items." Images are downsampled to ~225px thumbnails to keep the response small; the listing URL points to the full-size eBay page. Note: `Watchers` may read "less than 10 (eBay only reports watcher counts of 10+)" — that's not missing data, it's eBay's published-data threshold. Growth plan or higher.