BotSharper
任意のpump.funウォレットまたはトークン（Solana）に関する事実を、オンチェーンデータから提供。無料のリモートMCPサーバー、キー不要。botsharper_releve：公開Solanaアドレスが、1,300万件のクローズ済みポジションでスコアリングされた463,434人のpump.funオペレーターの中でどの順位か（パーセンタイル、五分位、勝ちポジションの割合、総損失の割合、平均倍率）。botsharper_jeton：pump.funトークンでローンチ以降に何が起きたか（ローンチブロック購入者、最初の1秒間の購入者、作成者の購入と売却、移行、購入者ランク）。botsharper_reperes：五分位ベンチマーク。過去の事実のみ、予測なし、アドバイスなし。
ホスト型 MCP サーバー
npx add-mcp 'https://botsharper.com/mcp'Claude Code、Codex、Cursor などにインストールできます
ドキュメント
Where do you actually stand.
A position report, computed across 463 434 real operators and 13.0 million closed positions. You paste a public address, you get its rank. Nothing else.
Public address only. Never a private key, never a recovery phrase: this site will never ask for one, and no serious site does.
See an example · Where do I find my address?
What the report actually says
01
A rank, not a score
Your position among every operator measured, across three separate indicators. A rank can be compared; a score can be argued with.
02
The rank holds over time
Measured across three separate periods: whoever is ahead stays ahead. Checked against a randomised control draw.
03
But it does not transfer
The measured edge comes from execution, not from picking assets. Copying someone's moves does not copy their rank.
What nobody else will tell you
Measurement, not opinion
No group of operators, not even the best fifth, clears break-even over the period measured. The rank exists and it is stable. Average profitability is not. A site that sells you the opposite is selling you a story.
What nearly fooled us
An early version of our measurements relied on a table whose total losses had been filtered out at build time. It reported a flattering average gain. Corrected, it reports the opposite. We published the correction rather than the flattering figure.
What we do not do
We hold no funds, no keys, no orders. We take no positions, we recommend no asset, and we earn nothing on whatever you do next with this report.
The dataset
463 434 operators measured
13.0 M closed positions
3 independent periods
The database is rebuilt in full every night from the complete sources, never updated by appending. Seven automatic checks block the publication of an inconsistent database.