strategy-lab
Stress-test lowcap, perp and LP strategies. Returns the full distribution and a reproduction seed.
使うべきか
品質と安全性
ツール定義とプロトコルへの準拠に関する自動分析に基づいています。
コンテキストコスト
これは、サーバーのツールがモデルのコンテキストに読み込まれるたびに消費されるおおよそのトークン数です。数が多いほど、ほかのタスクに使える注意が減ります。
インストール
ワンクリックインストール
これを `claude_desktop_config.json` ファイルに追加してください:
{
"mcpServers": {
"strategy-lab": {
"url": "https://kellyruns.xyz/mcp"
}
}
}リモートエンドポイント
https://kellyruns.xyz/mcpstreamable-httpできること
ツール一覧
ツール（5）
🟢get_agent_state
Kelly's current treasury, open positions, recent trades and the reasoning behind the latest cycle. Execution is in paper mode, so positions are simulated — treat this as a published track record, not a portfolio.
入力スキーマ
{
"type": "object",
"properties": {},
"additionalProperties": false
}🔴get_certification_gate
The exact thresholds a strategy must clear to be certified, and what a pass does and does not mean. Read this before interpreting any certified strategy: a certification is a statement about a distribution, not a prediction.
入力スキーマ
{
"type": "object",
"properties": {},
"additionalProperties": false
}🟢get_lab_summary
Counts by status, the testing pool balance, how many runs it can still fund, and the cost per run.
入力スキーマ
{
"type": "object",
"properties": {},
"additionalProperties": false
}🟢list_strategies(kind, status, sort, limit)
Strategies submitted to the lab, filterable by kind and status. Returns the certification rollup for each — run count, total paths, median PnL, worst-case PnL and ruin rate.
入力スキーマ
{
"type": "object",
"properties": {
"kind": {
"type": "string",
"enum": [
"lp",
"perp",
"lowcap"
],
"description": "Strategy domain."
},
"status": {
"type": "string",
"enum": [
"queued",
"testing",
"certified",
"rejected",
"inconclusive"
],
"description": "Lifecycle status."
},
"sort": {
"type": "string",
"enum": [
"top",
"new",
"queue"
],
"default": "top"
},
"limit": {
"type": "integer",
"minimum": 1,
"maximum": 100,
"default": 25
}
},
"additionalProperties": false
}🟢get_strategy(id)
Full detail for a single strategy: parameters, thesis, verdict, every check the gate ran, and each test run with its PRNG seed. The seed is the point — re-running with it reproduces the result exactly, so the claim is falsifiable.
入力スキーマ
{
"type": "object",
"properties": {
"id": {
"type": "string",
"description": "Strategy id."
}
},
"required": [
"id"
],
"additionalProperties": false
}コミュニティ
エビデンス