The full upstream README, mirrored here for reference. Install config, tool schemas, adoption signals, and an original overview live on the Earn Dataset MCP listing page.
MCP server over a static snapshot of Superteam Earn (superteam.fun) bounty listings: 28 open listings with their full description text — sponsor briefs, submission requirements, judging criteria — harvested 2026-09-04.
The data ships inside the package. No network calls at runtime, so the tools respond instantly and never fail on an upstream outage.
| Tool | What it does |
|---|---|
list_listings() | Index of all 28 listings: title, reward, token, deadline, URL |
get_listing(id_or_title) | One listing's complete description text. Accepts slug, exact title, or a unique title fragment |
search_listings(query) | Keyword search over titles + full descriptions, ranked by term frequency |
stats() | Dataset shape: count, harvest date, total reward pool ($56,705 face value), token mix, soonest deadlines |
Dataset fields per listing: slug, title, reward, token, deadline, desc. Rewards are in the listed token (15 USDG, 13 USDC); 2 listings have no stated reward.
If you want live data instead of a snapshot, see our other server, earn-bounty-scanner, which fetches from superteam.fun directly.
Running on FiatDock at $0.01/call: https://fiatdock.com/s/svc_7db4548f-dbf5-4f27-8e3c-31566b78ee01 (Streamable HTTP, pay-per-call via x402).
stdio:
Once the npm package is published, npx -y earn-dataset-mcp works too.
Snapshot pulled from live superteam.fun JSON endpoints on 2026-09-04, including the undocumented /api/search/{title} trick that returns full description text (the public cards API only exposes truncated fields). earn-api-endpoints.txt in the source repo maps 186+ reverse-engineered superteam.fun API routes if you want to harvest your own.
MIT