llmtrim
Level BLocalLocal proxy that compresses your LLM API requests so you pay less, with no change to the answers. Trims wasted tokens from prompts, history, tool output, and code before they're sent: -31% input / -74% output, measured live. Any provider, no extra model calls. Also an MCP server and embeddable library (Rust, Python, Ruby, Kotlin, Swift, JS/TS).
Installation
MCP client config (paste into mcp.json)
{
"mcpServers": {
"llmtrim": {
"command": "npx",
"args": [
"@llmtrim/cli"
]
}
}
}Agent install prompt
Copy and paste into Cursor / Claude / other agents to install this MCP for you
Please read this MCP repository and install it into my current environment:
https://github.com/fkiene/llmtrim
Known working client config (can write into mcp.json):
{
"mcpServers": {
"llmtrim": {
"command": "npx",
"args": [
"@llmtrim/cli"
]
}
}
}
Please: 1) check README for deps and env vars; 2) write config for my client; 3) if secrets are missing, tell me what to fill — do not invent values.
Related servers
REA逆向工程一体化 MCP:把 Ghidra、IDA、Hopper、JADX、Frida、ADB、浏览器等工具统一封装,Agent 无需源码即可分析二进制、App、gemini-cliAn open-source AI agent that brings the power of Gemini directly into your termiawesome-mcp-serversA collection of MCP servers.clineAutonomous coding agent as an SDK, IDE extension, or CLI assistant.Context7为 AI 提供最新版技术文档和代码示例检索,写代码不再用过时 API,免密钥开箱即用next-ai-draw-ioDraw and edit draw.io diagrams by chatting with an AI. Web app, desktop app, and