NewCost control: a token allowance a month for each model choice — when one runs out, the cheaper one answers.See how it works
memtro
Client

Memtro for Cline

Choose OpenAI Compatible as Cline's provider with Memtro's base URL, and add Memtro as a remote MCP server.

Cline runs in VS Code and takes any OpenAI-compatible endpoint: set the provider to OpenAI Compatible, the base URL to https://platform.memtro.com/v1 and the model to auto, and every task runs with your organisation's memory and routing behind it. Add Memtro under Remote Servers and the agent can read your systems while it works.

What you get

  • OpenAI Compatible provider: base URL https://platform.memtro.com/v1, model auto, memtro or any provider/model
  • Remote MCP server over Streamable HTTP with memory and connector tools, per-tool auto-approve
  • Tasks that check the CRM, the ticket or the ledger before touching code

How to connect

  1. 01Cline settings (⚙) → API Provider: OpenAI Compatible → Base URL https://platform.memtro.com/v1, API Key sk-memtro-…, Model auto
  2. 02MCP Servers icon → Remote Servers → Server Name memtro, Server URL https://platform.memtro.com/api/mcp, Transport Type Streamable HTTP → Add Server
  3. 03Then add the key to the memtro entry in ~/.cline/data/settings/cline_mcp_settings.json: "headers": { "Authorization": "Bearer sk-memtro-…" }

Example queries

  • › Before changing the invoice email, check how Help Scout customers describe the problem
  • › auto:agent as the model for tasks that need the systems checked first

Why it is better with memory

Through Cline you get the same memory, connector tools, routing and jobs as every other client, so what one person learns in one tool is available to the whole team in all of them.