MCP.so
ログイン
B

Budget Governor

@mightbesaad

リポジトリ

Budget Governor について

Budget & cost control for AI agents: hard per-agent spend caps, rate limits, idempotency, and human-in-the-loop approval — enforced before each LLM call, not after the invoice. One hosted MCP endpoint (no proxy or self-hosting), settled via x402 (USDC on Base).

#budget#cost-control#spend-cap#rate-limit#ai-agents#idempotency#x402#governance

基本情報

カテゴリ

AI とエージェント

トランスポート

stdio

公開者

mightbesaad

投稿者

mightbesaad

設定

以下の設定を使って、このサーバーを MCP 対応クライアントに追加してください。

{
  "mcpServers": {
    "gvnr": {
      "url": "https://gvnr.dev/mcp",
      "headers": {
        "Authorization": "Bearer YOUR_GVNR_API_KEY"
      }
    }
  }
}

ツール

ツールは検出されませんでした

ツールは README から自動的に抽出されます。メンテナーは ## Tools という見出しの下に記載することで、このタブに反映できます。

概要

What is Budget Governor?

Budget Governor is a hosted MCP server that enforces spend caps, rate coordination, idempotency, and a human‑in‑the‑loop gate for AI agents – enforced before the LLM call, not after the invoice. It integrates via one MCP endpoint and settles using x402 (USDC on Base). It is built for developers running AI agents who need cost controls without self‑hosting proxy infrastructure.

How to use Budget Governor?

Add the MCP URL https://gvnr.dev/mcp?api_key=bg_YOUR_KEY to any MCP‑compatible client. Provision an account via POST /v1/account (includes 25 free trial ops), top up your governance‑operation quota with USDC, set a spend envelope per agent, then call budget_clear before each LLM request and reconcile after.

Key features of Budget Governor

  • Pre‑call spend clearance against a per‑agent USD envelope
  • Independent governance‑operation quota (1,000 ops per $1 USDC)
  • Per‑(agent, provider, model) rate limiting with retry‑after hint
  • Request‑level idempotency deduplication for retry safety
  • Human‑in‑the‑loop approval flow with polling API
  • No proxy – tokens never pass through the server

Use cases of Budget Governor

  • Prevent runaway agent costs by capping daily or per‑session spend
  • Coordinate request rates across multiple agents sharing a model provider
  • Deduplicate retried LLM calls to avoid double billing
  • Pause execution for human approval on high‑value or sensitive actions
  • Meter and bill agent usage with a pay‑as‑you‑go governance quota

FAQ from Budget Governor

What does Budget Governor NOT do?

Budget Governor is not an LLM proxy – your tokens never pass through it, and you pay your model provider directly. It governs only the decision to spend.

What are the runtime requirements?

No self‑hosting, infrastructure, or proxy to deploy. It is a hosted endpoint accessible via HTTP.

Where does data live?

All state (op quota, envelopes, approvals) is managed server‑side on gvnr.dev infrastructure. Your API key is the credential.

What transports and auth are supported?

HTTP with Bearer token (API key) for both REST and MCP. The MCP endpoint uses the URL https://gvnr.dev/mcp?api_key=bg_YOUR_KEY.

How does

コメント

「AI とエージェント」の他のコンテンツ

Sequential Thinking Multi-Agent System (MAS)

FradSer

An advanced sequential thinking process using a Multi-Agent System (MAS) built with the Agno framework and served via MCP.

306

MCP Manager for Claude Desktop

zueai

simple web ui to manage mcp (model context protocol) servers in the claude app

290

Gemini MCP Server

aliargun

MCP server implementation for Google's Gemini API

259

1Panel

1Panel-dev

🔥 1Panel is a modern, open-source VPS control panel — and the only one with native AI agent support. Run Ollama models, deploy OpenClaw agents, and manage your entire server stack from one clean web interface.

36.1K

Perplexity MCP Server

DaInfernalCoder

A Model Context Protocol (MCP) server for research and documentation assistance using Perplexity AI. Won 1st @ Cline Hackathon

295