Weckr MCP server

What it does

Weckr is an MCP server that tracks AI cost and margin intelligence for SaaS applications. It integrates with your LLM calls to measure which customers cost more in API fees than they pay, broken down by user, feature, and model, then surfaces this data through a dashboard with recommendations for cost optimization.

• Identify which paying customers are unprofitable due to high LLM API costs and adjust pricing or usage limits accordingly.
• Compare per-token costs across different LLM providers and models to find cheaper alternatives without changing your application logic.
• Set per-plan spending caps that the SDK enforces before making LLM calls to prevent runaway costs for specific customer tiers.

Tools this server exposes

Tool	Does
weckr_track_llm_call	Wrap an LLM API call to capture cost metadata (model, tokens, latency, user ID, feature, plan) and send it to Weckr for analysis.
weckr_get_dashboard_data	Retrieve aggregated cost and margin data per user, feature, or model from the Weckr dashboard.
weckr_check_spending_cap	Verify whether a customer or plan has exceeded its spending cap before executing an LLM call.

Before you enable it

An MCP server runs with whatever access you give it. Start with a development account rather than production, and check which tools the server exposes before you let an agent call them unattended.