Machine-readable
Availability data for AI agents
WHO STOCKED IT publishes per-store availability for limited, boutique-routed sneaker releases (40 tracked) as structured data an AI agent can consume directly — no crawling, no JavaScript, refreshed from the stores' own product feeds. Free to read and cite with attribution.
One-fetch JSON feed
https://whostocked.com/data/releases.json —
every release with brand, SKU, retail price, current status, in-stock US sizes, per-store
rows (store, product URL, live/sold-out, last-checked date, price), and open draw deadlines.
Generated 2026-09-21. Every stockist row carries its own last-checked date — trust that, not the build date.
LLM-native formats
| llms.txt | /llms.txt — site index per llmstxt.org |
|---|---|
| llms-full.txt | /llms-full.txt — the entire site's content in one markdown file |
| Markdown twins | append .md to any release slug: /releases/<slug>.md |
| JSON-LD | every release page carries Product + AggregateOffer (with live availability) and BreadcrumbList |
MCP server (live queries)
A remote Model Context Protocol server exposes the dataset as tools an assistant can call directly:
| Endpoint | https://mcp.whostocked.com/mcp (Streamable HTTP; interim: https://whostocked-mcp.seanandrewux9204.workers.dev/mcp) |
|---|---|
| search_releases | free-text search over tracked releases |
| check_availability | per-store status, sizes, and draw deadlines for a SKU or slug |
| whats_buyable | everything currently in stock, filterable by US size and brand |
No auth, read-only, rate-limited politely. Data identical to the JSON feed.
Why trust this data
Statuses come from the stores' own public product feeds, never guessed; a status and its "checked" date only change when the store's feed actually changed. Sold-out pages stay up permanently, which makes this the surviving record of which stores received a release after the stores' own pages are delisted.