DEV Community: publicrecords

The MCP servers an Etsy seller can actually use

publicrecords — Tue, 03 Nov 2026 01:29:52 +0000

Search "Etsy MCP" and you'll find a handful of half-finished experiments mixed in with general scraping tools. I wanted a short, honest list of what actually works for an Etsy seller who uses Claude, Cursor or ChatGPT, so I made one: Awesome Etsy MCP.

How it's organized

One entry per data source or marketplace. Each one was checked when it was added.

Etsy itself

Etsy Dev MCP Server is Etsy's own server. It searches the Open API v3 docs and schemas while you build. It doesn't call the API for you.
etsy-mcp-server (community) exposes the Open API v3 for your own shop: listings, inventory, orders, shipping and shop settings, with the OAuth scope each call needs.
Etsy Search & Shop Sales MCP is ours and is marked that way. It returns live Etsy search rows (title, price, shop, badges, rank) and daily sales counters for competitor shops you name.

Scraping platforms that cover Etsy

Apify, Bright Data and Firecrawl all have official MCP servers that can read public Etsy pages alongside other stores. They're the right pick if you need Etsy plus everything else.

Other marketplaces

Most sellers aren't only on Etsy, so the list has one entry each for Amazon (the official Selling Partner connector), eBay (Sell APIs) and Shopify (dev docs and schemas).

Things you can ask once it's connected

"What are the top 20 Etsy results for 'ceramic planter' right now, and which shops show up more than once?"
"Which of these three competitor shops is growing fastest this month?"
"Draft a listing title in the style of the best-ranked results for this keyword."

Connecting

With Claude Code:

claude mcp add --transport http etsy-dev "https://mcp.api.etsycloud.com/mcp"

The README has Cursor and VS Code snippets for every entry.

Suggest one

The list is CC0 and open to pull requests. If something covers a source that isn't listed, or clearly beats a current entry, send it in.

Disclosure: I maintain the list and one entry on it. The rest are independent. Use these tools only on data you're allowed to access, and follow each marketplace's terms.

The sports odds MCP servers worth connecting, one per sportsbook

publicrecords — Tue, 03 Nov 2026 01:29:32 +0000

If you've tried giving Claude, Cursor or ChatGPT live betting odds, you've probably noticed that "sports odds MCP" search results are mostly duplicates: several wrappers around the same aggregator, abandoned demos, and a few tools that scrape logged-in sessions. Picking one takes longer than it should.

So I put together a short curated list: Awesome Sports Odds MCP.

The one rule: one provider per sportsbook

Each sportsbook, book network, exchange or aggregator gets a single entry, the strongest one I could find. If two servers cover the same source, only the better one is listed. That keeps it short enough to read in a couple of minutes.

What's on it

Multi-book aggregators. These normalize many books into one schema: OpticOdds, Odds-API.io, SportsGameOdds, TheRundown and PropLine (for player props).

Sportsbook networks and single books. These read one source directly. Our own Kambi server (BetRivers, Unibet, Paf, Svenska Spel and ATG) is here, marked as ours, next to PinnWire for Pinnacle and OddiWire for esports.

Regional bookmakers. PuntersEdge covers Australian racing and sports.

Prediction markets. One Polymarket server and one Kalshi server, for anyone pricing event contracts against sportsbook lines.

Docs MCPs for builders. Sportradar and OddsPapi publish MCP servers that serve their API docs and schemas instead of prices. They're useful while you're writing an integration.

What I left out

Servers with no recent activity
Near-duplicates of an entry already on the list
Anything that depends on borrowed cookies or logged-in sessions
Anything selling win-rate or "guaranteed edge" claims

Connecting one

Most entries are remote MCP servers, so you just add a URL. With Claude Code:

claude mcp add --transport http sports-odds "https://mcp.apify.com?tools=publicrecords/sports-odds-api"

Or in Cursor (~/.cursor/mcp.json):

{
  "mcpServers": {
    "pinnacle": { "url": "https://pinnwire.com/mcp" }
  }
}

Each provider's own docs explain whether it expects an API key or OAuth.

Help keep it current

The list is CC0. If you know a server that covers a book that isn't listed yet, or one that beats a current entry, open a pull request. The contributing guide covers the one-per-sportsbook rule.

Disclosure: I maintain the list and one of its entries (the Kambi server). Everything else is independent and listed on merit. Odds data only; follow the laws where you live.

Track Etsy Shop Sales Velocity with an Apify Actor (curl + Python)

publicrecords — Tue, 27 Oct 2026 15:00:22 +0000

Track Etsy Shop Sales Velocity with an Apify Actor (curl + Python)

Disclosure: I maintain this Actor.

If you sell on Etsy (or research niches), you eventually want a simple question answered: how fast is that other shop selling? Etsy's public shop page shows a lifetime sales counter. Read it twice, with dates, and you can derive a measured sales-per-day rate. This tutorial walks through a real run of the public Etsy Shop Sales Tracker & Scraper Actor on Apify — not a mock.

What you get back

One dataset row per shop. Useful fields include:

shop, shop_url, title, category
sales_count — the public lifetime sales counter as read
as_of / snapshot_date — when that counter was read
sales_per_day — rate derived from consecutive exact reads
delta_7d, delta_last, history_days
public engagement counters: reviews_count, rating, admirers, listings_active

The Actor tracks shops it can read on the public page. It is not limited to Etsy Bestsellers, and this post does not claim an "only Bestsellers" scope.

Pricing

from $0.00005 per run start + $0.003 per shop row (free plan, lower on paid plans), as of Oct 27, 2026. Re-check the Store pricing page before budgeting a large crawl.

A tested run

On Oct 27, 2026 I ran the Actor with the Store example input, {"shops": ["OrelCeramics", "Bestickend"]}. It succeeded in a few seconds and returned 2 shop rows.

The two rows it returned:

shop	sales_count	sales_per_day	delta_7d	delta_last	as_of	history_days
OrelCeramics	111	0	0	0	2026-10-26	27
Bestickend	11,475	2	10	6	2026-10-26	27

Bestickend shows a clear weekly change (delta_7d = 10) and a recent step (delta_last = 6) against a dated counter (as_of = 2026-10-26). OrelCeramics is quieter on the same window — useful when you want a side-by-side contrast, not only hot shops.

Call it with curl

Replace APIFY_TOKEN with your Apify API token (Console → Settings → Integrations). Do not commit the real token.

1. Start a run (Store example input)

curl -sS -X POST \
  'https://api.apify.com/v2/acts/publicrecords~etsy-shop-velocity/runs?token=APIFY_TOKEN' \
  -H 'Content-Type: application/json' \
  -d '{
    "shops": ["OrelCeramics", "Bestickend"]
  }'

Useful input fields (from the Store input schema):

shops — array of shop names (the path after etsy.com/shop/)
keywords — match title / headline
category — panel category filter
minSales — lifetime sales floor
since — ISO date; only shops whose counters changed since then
maxShops — upper bound on rows returned (and charged)

2. Wait for the run, then read the default dataset

# RUN_ID and DATASET_ID come from the start response:
# .data.id and .data.defaultDatasetId
curl -sS \
  "https://api.apify.com/v2/actor-runs/RUN_ID?token=APIFY_TOKEN" \
  | jq '{status: .data.status, datasetId: .data.defaultDatasetId}'

# rows, once status is SUCCEEDED
curl -sS \
  "https://api.apify.com/v2/datasets/DATASET_ID/items?token=APIFY_TOKEN&clean=1&limit=10" \
  | jq '.[] | {shop, sales_count, sales_per_day, delta_7d, delta_last, as_of}'

3. Schedule it

In the Apify Console, open the Actor → Schedules, or create a schedule via the Schedules API pointing at the same Actor. Daily or weekly reads are enough to keep sales_per_day and delta_7d meaningful; denser schedules mostly burn row events.

How to read the numbers

sales_count + as_of — an exact public counter at a known date. Treat it as a measurement, not an Etsy Analytics export.
sales_per_day — the shop is read a few days apart; sales per day is the average over that gap. If history is short, the rate is noisier.
delta_7d / delta_last — change across roughly a week, and the most recent step, when the history supports it; otherwise check null-reason fields on the row.
Compare shops side-by-side on sales_per_day and delta_7d, not on lifetime sales_count alone (a huge shop can grow slowly; a small shop can be accelerating).

Minimal Python client

import os, time, requests

TOKEN = os.environ["APIFY_TOKEN"]  # never hard-code
ACT = "publicrecords~etsy-shop-velocity"
base = "https://api.apify.com/v2"

r = requests.post(
    f"{base}/acts/{ACT}/runs",
    params={"token": TOKEN},
    json={"shops": ["OrelCeramics", "Bestickend"]},
    timeout=60,
)
r.raise_for_status()
run = r.json()["data"]
run_id = run["id"]

while True:
    st = requests.get(f"{base}/actor-runs/{run_id}", params={"token": TOKEN}, timeout=60)
    st.raise_for_status()
    data = st.json()["data"]
    if data["status"] in {"SUCCEEDED", "FAILED", "ABORTED", "TIMED-OUT"}:
        break
    time.sleep(2)

ds = data["defaultDatasetId"]
items = requests.get(
    f"{base}/datasets/{ds}/items",
    params={"token": TOKEN, "clean": 1},
    timeout=60,
).json()
for row in items:
    print(
        row["shop"],
        row.get("sales_count"),
        row.get("sales_per_day"),
        row.get("delta_7d"),
        row.get("as_of"),
    )

Where this fits

Competitor shortlists: pin 10–50 shops, schedule weekly, watch delta_7d.
Niche scans: pass keywords / category instead of (or with) explicit shop names.
Free public slice: etsypulse.ca publishes movers/rising from the same family of reads; each shop row there links back to this Actor via "Track this shop's sales".

Limits to say out loud

Public counters only — not order-level or revenue data.
A shop can hide sold listings; the lifetime counter may still be readable when listings are not.
Pricing and field labels can change; re-read the Store page before a large job.

Disclosure: I maintain this Actor.

Store page: https://publicrecords-redirect.publicrecords.workers.dev/r/devto-etsy-1

Find bestselling Etsy listings in a niche: 10 rows for $0.04

publicrecords — Mon, 05 Oct 2026 20:16:23 +0000

Disclosure: I maintain these Actors on Apify Store (publicrecords).

Etsy sellers and product researchers keep asking how to export Etsy search results for a niche and see which listings are already marked Bestseller. Today's walkthrough does that with Etsy Search Scraper — a different path from sorting by price.

Price (matches the live listing)

Pay-per-event on Apify Store:

Actor start: $0.00005
Listing row: $0.004 per delivered listing (free plan; lower on paid plans, down to $0.0034)
10 rows: $0.00005 + 10 × $0.004 = $0.04005 (about $0.04)

Prices as of Oct 27, 2026 — see the Store pricing page.

Blocked pages are not charged.

Paste-ready input (bestsellers in a niche)

{
  "queries": ["personalized dog collar"],
  "sort": "relevance",
  "is_best_seller": true,
  "maxItems": 10,
  "maxPages": 1,
  "fillLazyCards": true,
  "proxyConfiguration": {
    "useApifyProxy": true,
    "apifyProxyGroups": ["RESIDENTIAL"],
    "apifyProxyCountry": "US"
  }
}

Swap the keyword for your niche. is_best_seller: true keeps only listings Etsy marks Bestseller.

Run it

Open the Actor: publicrecords/etsy-search-scraper
Paste the JSON into Input → Start
Download the dataset (CSV/JSON)

Or from the CLI:

apify call publicrecords/etsy-search-scraper --input '{
  "queries": ["personalized dog collar"],
  "sort": "relevance",
  "is_best_seller": true,
  "maxItems": 10,
  "maxPages": 1,
  "fillLazyCards": true,
  "proxyConfiguration": {
    "useApifyProxy": true,
    "apifyProxyGroups": ["RESIDENTIAL"],
    "apifyProxyCountry": "US"
  }
}'

curl (replace $APIFY_TOKEN):

curl -sS -X POST "https://api.apify.com/v2/acts/publicrecords~etsy-search-scraper/runs?token=$APIFY_TOKEN" \
  -H 'Content-Type: application/json' \
  -d '{
    "queries": ["personalized dog collar"],
    "sort": "relevance",
    "is_best_seller": true,
    "maxItems": 10,
    "maxPages": 1,
    "fillLazyCards": true,
    "proxyConfiguration": {
      "useApifyProxy": true,
      "apifyProxyGroups": ["RESIDENTIAL"],
      "apifyProxyCountry": "US"
    }
  }'

What you get

Each row is a live search listing: id, title, price, rating, reviews, badges (including Bestseller), shop, and rank in Etsy's order for that query.

Example repo for the price-sort path: GeminiGeorge22/etsy-search-example.

Again: disclosure — I maintain publicrecords/etsy-search-scraper on Apify Store.

Scrape Etsy search results sorted by price: 10 rows for $0.04

publicrecords — Mon, 05 Oct 2026 17:58:37 +0000

Disclosure: I publish and maintain the Actor used here (publicrecords/etsy-search-scraper on Apify Store). This is the publisher account.

You have an Etsy search query and you want the results as rows you can sort and filter, not a screenshot of page one. Here's the shortest route I know, with a real run as proof.

What you get

Each Etsy search result comes back as one row with listing_id, title, price, currency, shop_name, shop_url, rating_value, review_count, badges (bestseller, star_seller, etsys_pick, popular_now, free_shipping), position and page. You can pass a keyword, a market, or a category URL.

The input (copy this)

This is the exact input from a Store-page run on 2026-10-03: the 10 cheapest "ceramic mug" listings priced $10–$40.

{
  "queries": ["ceramic mug"],
  "sort": "price_asc",
  "minPrice": 10,
  "maxPrice": 40,
  "maxItems": 10,
  "maxPages": 1,
  "fillLazyCards": true,
  "proxyConfiguration": {
    "useApifyProxy": true,
    "apifyProxyGroups": ["RESIDENTIAL"],
    "apifyProxyCountry": "US"
  }
}

sort takes one of relevance, newest, price_asc, price_desc, top_reviews, the same options as Etsy's own sort menu.
minPrice / maxPrice are whole currency units.
maxItems and maxPages cap the run (Etsy itself stops at 20 pages).

Run it

In the browser: open the Actor in Apify Console, switch the input to JSON, paste the block above, and press Start.

From a terminal: save the block as input.json and run:

curl -s -X POST \
  "https://api.apify.com/v2/acts/publicrecords~etsy-search-scraper/run-sync-get-dataset-items?token=$APIFY_TOKEN" \
  -H "Content-Type: application/json" \
  -d @input.json > rows.json

Or clone the one-command example repo, which has the same input.json and the full sample output: GeminiGeorge22/etsy-search-example.

Proof run

Field	Value
Actor	publicrecords/etsy-search-scraper
Run	Started from the Store page, 2026-10-03
Status	SUCCEEDED, 10 rows
Cost (current price)	actor-start $0.00005 + 10 × listing-row $0.004 = $0.04005 (about $0.04), free plan, as of Oct 27, 2026 — pricing

The first 3 of the 10 rows:

#	Title	Price	Shop	Rating	Reviews
1	1980's Christmas Coffee Mug	$10	TheNostalgiaHunter	4.9	~2,200*
2	Daisy a Day Coffee Mug, 11oz Ceramic Mug	$10	SmugMugMugs	5.0	51
3	America 250th Birthday Mug, 1776 Bicentennial Style	$10	SouthernRoyaltyUS	5.0	32

*Etsy shows "2.2k", so that row comes back with review_count_approx: true instead of a made-up exact number.

Pricing, plainly

$4 per 1,000 listing rows on the free plan (lower on paid plans, down to $3.40), plus a $0.00005 start fee per run — as of Oct 27, 2026; see the Store pricing page.
Runs on Apify residential proxy by default. That's passthrough: Apify bills proxy usage, and I add no markup.
A blocked page costs nothing. If a row comes back without a price, title or shop, you still get it, flagged incomplete: true, and you aren't charged for it.

Next step: who is selling?

Search rows tell you what ranks. If you also want to know how fast a shop is selling, feed the shop_name values into the companion Actor Etsy Shop Sales Tracker. It returns one row per shop with the public lifetime sales counter and the date it was read. The shop is read a few days apart; sales per day is the average over that gap.

Questions or a query that breaks it? Leave a comment. I maintain both Actors and I read the comments.