mdfifty50-boop avatar

Agent Costcenter

mdfifty50-boop/agent-costcenter-mcp

If you're running multiple AI agents and need to understand where your API costs are going, this server gives you a cost center for your fleet. It tracks spend per agent, team, or project by logging LLM and tool calls against a built-in pricing table covering Claude, GPT, Gemini, and DeepSeek models. You can set budget caps with warn or block thresholds, pull cost breakdowns showing model-level detail and percentages, and surface anomalies when an agent's spending spikes above 2x normal. The fuzzy model name matching means you don't need exact strings, and the compare_models tool shows you cost per 1K tokens across everything you're using. Useful when you need attribution and guardrails across a multi-agent system.

tools
8
transport
STDIO
last updated
Apr 24, 2026

Tools

8 tools

Public tool metadata for what this MCP can expose to an agent.

register_agentRegister agent with team, project, model, and budget cap

Register agent with team, project, model, and budget cap

No parameter schema in public metadata yet.

log_llm_callLog an LLM call — auto-calculates cost from pricing table

Log an LLM call — auto-calculates cost from pricing table

No parameter schema in public metadata yet.

log_tool_callLog a tool/API call cost

Log a tool/API call cost

No parameter schema in public metadata yet.

get_cost_reportCost breakdown by agent, team, project, or all

Cost breakdown by agent, team, project, or all

No parameter schema in public metadata yet.

set_budget_alertSet spending threshold (warn or block)

Set spending threshold (warn or block)

No parameter schema in public metadata yet.

check_budgetCheck if agent is within budget

Check if agent is within budget

No parameter schema in public metadata yet.

get_cost_anomaliesFind agents with unusual spending spikes

Find agents with unusual spending spikes

No parameter schema in public metadata yet.

compare_modelsCompare cost efficiency across models

Compare cost efficiency across models

No parameter schema in public metadata yet.

Sponsors

Readme

agent-costcenter-mcp

MCP server for per-agent cost attribution and budget management. Answer "which agent costs the most?" across your entire AI agent fleet.

Features

  • Per-agent cost tracking — register agents, log LLM and tool calls, get per-agent spend
  • Built-in pricing table — 11 models (Claude, GPT, Gemini, DeepSeek) with fuzzy name matching
  • Budget management — set caps and alerts, check remaining budget, warn or block on overspend
  • Cost reports — breakdown by agent, team, project, or all; includes model-level detail and percentages
  • Anomaly detection — automatically flag agents with 2x+ spending spikes
  • Model comparison — compare cost-per-1K-tokens across all models used

Quick Start

npx agent-costcenter-mcp

Claude Desktop

{
  "mcpServers": {
    "agent-costcenter": {
      "command": "npx",
      "args": ["agent-costcenter-mcp"]
    }
  }
}

Tools


View the full README on GitHub

Details

Repository
github.com/mdfifty50-boop/agent-costcenter-mcp
Categories
AI & LLM Tools, Finance & Commerce
Package
agent-costcenter-mcp
Transport
STDIO
Registry
MCP registry, active

More from mdfifty50-boop

Related AI & LLM Tools MCP servers

View all
1mikerawsonnz avatarTraced Llm Proxymikerawsonnz/gosce-agentsProxy Gemini (Vertex AI) completions wrapped in OpenTelemetry trace spans; returns the answer plus t0stars
2mdvohra avataragent-godmodemdvohra/agent-godmodeSandboxed workspace MCP: file ops, argv-only run_command, list_files; BYO LLM prompts.0stars
3weiseer avatarllm-oracleweiseer/llm-oracleUp-to-date LLM pricing + availability catalog with cited sources for AI agent routing.0stars
4builderio avatarAgent-Native Analyticsbuilderio/agent-nativeAgent-Native Amplitude/Mixpanel - connect data sources, prompt for charts7.6Kstars
5builderio avatarAgent-Native Dispatchbuilderio/agent-nativeCentral Slack/Telegram router with jobs, memory, approvals, and A2A delegation7.6Kstars
6assafelovic avatarGPT Researcherassafelovic/gptr-mcpMCP server for enabling LLM applications to perform deep research via the MCP protocol346stars