Poshmark Comps MCP server

@pipeworx/poshmark-comps

Sold-price and active-listing search on the public Poshmark marketplace — the best open sold-price comp data on Poshmark, with no login and no API key.

Part of Pipeworx — an MCP gateway connecting AI agents to 1576+ live data sources.

Tools

listings, returns price, size, brand, condition, sold date, seller, and a listing URL per hit.

• poshmarksoldsearch(query, department?, sort_by?, limit?) — search SOLD

for currently-listed (not yet sold) items, for comparing ask vs. sold.

• poshmarkactivesearch(query, department?, sort_by?, limit?) — same shape

Auth

Keyless. No login, no API key, no anti-bot challenge encountered (verified live 2026-09-05).

Data sources

no separate JSON API endpoint: the page is server-rendered by a Vue app and embeds the full result set in a window.INITIALSTATE__ script-tag blob (checked live: an XHR/Accept: application/json request returns the same full HTML page, not JSON). This pack fetches that HTML and extracts the embedded JSON directly — cheaper and more robust than DOM parsing.

• — Poshmark's own public search page. There is

Traps for the next person:

nextmaxid in the response is a real pagination cursor but this pack doesn't walk it yet — totalmatchesreportedbyposhmark in the tool output tells the caller how much more exists.

• One page per call, up to ~48 results (Poshmark's own default page size).

post (what the seller was asking), not a separately verified receipt/hammer amount, and excludes Poshmark's fee and shipping. State this to callers — it's in coverage_note on every response.

• price on a sold item is the listing price Poshmark shows on the sold

priceasc — newest and mostloved returned empty result sets when probed live 2026-09-05; not exposed as options for that reason.

• sortby only reliably takes bestmatch (default), price_desc,

(results included non-matching brands); free-text query does the real filtering, so brand/size aren't exposed as separate parameters — put them in query instead (e.g. "lululemon align size 6").

• brand/size query params on the site did not reliably filter when probed

concurrent fan-out.

• Crawl politely: one page fetch per tool call, no pagination loop, no

Quick Start

Add to your MCP client (Claude Desktop, Cursor, Windsurf, etc.):

{
  "mcpServers": {
    "poshmark-comps": {
      "url": "https://gateway.pipeworx.io/poshmark-comps/mcp"
    }
  }
}

What this endpoint actually serves

tools/list at https://gateway.pipeworx.io/poshmark-comps/mcp returns the tools in the table above plus the shared Pipeworx meta-tools — askpipeworx, discovertools, search_within, remember/recall and the rest of the gateway-wide set. So the tool count you see is larger than this table: a single-pack endpoint currently lists roughly 30 shared tools alongside the pack's own. The connection's initialize response states its exact scope, and is the authoritative answer for a given day.

This is deliberate, not multiplexing by accident. The meta-tools are what let a scoped connection answer a question this pack does not cover — via ask_pipeworx, which routes across the whole catalog — without you adding a second MCP server. There is currently no way to mount a pack endpoint without them; if the extra schemas cost you more context than the routing is worth, connect to the full gateway once rather than to several pack endpoints.

Or connect to the full Pipeworx gateway to get every pack's tools listed directly, instead of just this one's:

{
  "mcpServers": {
    "pipeworx": {
      "url": "https://gateway.pipeworx.io/mcp"
    }
  }
}

Both URLs reach the same gateway and the same 1576+ data sources. The only difference is which pack's tools are listed directly; ask_pipeworx reaches all of them from either one.

Shortened. The full README is on GitHub.