LLM Brand Monitor

io.github.SerpstatGlobal由 GitHub 账号 SerpstatGlobal 发布v1.0.1更新于 Oct 24, 2026

Track how 350+ AI models mention your brand — manage projects, run scans, analyze results

已验证STDIO仅桌面数据与分析商业与电商AI 与机器学习

概览

AI 生成的概览

本地 MCP 服务器，让助手管理 LLM Brand Monitor 项目、在 350+ 个 AI 模型上运行品牌提及扫描并分析结果。

可以做什么
列出、查看、创建、更新和归档品牌监测项目，并为项目添加或删除监测提示词。启动扫描，将提示词发送给大模型，查看扫描进度和扫描历史。浏览结果、读取完整的模型回复、查看被提及的竞品和被引用的网址，并查看竞品提及趋势。列出可用的模型，查询积分余额和使用统计。
适用场景
适合已在使用 LLM Brand Monitor 的用户，希望通过对话完成项目与提示词设置、发起扫描，并查看 AI 模型如何提及自身品牌和竞品。
运行要求
通过 npx（Node.js）或从源码构建后本地运行。需要 LLM Brand Monitor 账号，并通过环境变量 LBM_API_KEY 提供 API 密钥。可选环境变量 LBM_API_BASE_URL 和 LOG_LEVEL。需要访问 LLM Brand Monitor API 的网络连接。
安装前请注意
LBM_API_KEY 是密钥，可访问你的账号。扫描会消耗积分，README 说明助手应在运行前请求确认；积分不足会返回错误。工具可创建、更新、归档和删除项目及提示词，会实际改动账号数据。提示词和结果会发送至 LLM Brand Monitor 服务。

安装

在 SourceWeft 中

  1. 打开 控制台中的 LLM Brand Monitor，将其添加到工作区。
  2. 为需要使用其工具的对话启用该服务。

Desktop only，通过 STDIO。 STDIO 服务会启动本地进程，因此需要 SourceWeft 桌面宿主。

其他 MCP 客户端

参照 仓库 中的启动说明。

README

LLM Brand Monitor MCP Server

[npm version] [MCP Badge] [License: MIT]

MCP server for LLM Brand Monitor — a platform that tracks how AI models mention your brand.

Connect Claude, Cursor, Windsurf, or any MCP-compatible client to manage brand monitoring projects, run scans across 350+ LLMs, and analyze results — all through natural language.

Quick Start

Claude Desktop

Add to your claude_desktop_config.json:

json
{  "mcpServers": {    "lbm": {      "command": "npx",      "args": ["-y", "@serpstat/llm-brand-monitor-mcp"],      "env": {        "LBM_API_KEY": "lbm_your_key_here"      }    }  }}

Claude Code

bash
claude mcp add lbm-mcp -e LBM_API_KEY=lbm_your_key_here -- npx -y @serpstat/llm-brand-monitor-mcp

From Source

bash
git clone https://github.com/SerpstatGlobal/llm-brand-monitor-mcp.gitcd llm-brand-monitor-mcpnpm install && npm run build

Then add to your MCP client config:

json
{  "mcpServers": {    "lbm": {      "command": "node",      "args": ["/path/to/llm-brand-monitor-mcp/dist/index.js"],      "env": {        "LBM_API_KEY": "lbm_your_key_here"      }    }  }}

MCP Inspector

bash
LBM_API_KEY=lbm_your_key_here npx @modelcontextprotocol/inspector node dist/index.js

Getting an API Key

  1. Sign up at llmbrandmonitor.com
  2. Open your Profile page
  3. Copy your API key (starts with lbm_)
[API key in LLM Brand Monitor profile]

Tools

17 tools across 4 categories:

Projects (7 tools)

ToolWhat it does
lbm_list_projectsList all brand monitoring projects
lbm_get_projectGet project details with prompts and models
lbm_create_projectCreate a new project
lbm_update_projectUpdate project name, models, or monitoring settings
lbm_archive_projectArchive a project
lbm_add_promptsAdd monitoring prompts to a project
lbm_delete_promptRemove a prompt from a project

Scans (3 tools)

ToolWhat it does
lbm_run_scanStart a scan — sends prompts to LLMs and collects responses
lbm_get_scan_statusCheck scan progress
lbm_list_scansView scan history

Results (5 tools)

ToolWhat it does
lbm_list_resultsBrowse monitoring results (brand mentions, status)
lbm_get_transcriptRead the full LLM response for a specific result
lbm_list_competitorsSee which competitor brands LLMs mention
lbm_list_linksSee which URLs and domains LLMs cite
lbm_get_historyCompetitor mention trends over time

Models & Usage (2 tools)

ToolWhat it does
lbm_list_modelsList 350+ available LLM models
lbm_get_usageCheck credit balance and usage stats

Typical Workflow

You: "What brand monitoring projects do I have?"Claude: → lbm_list_projects
You: "Run a scan on the Serpstat project"Claude: → lbm_run_scan (asks you to confirm — scans spend credits)       → lbm_get_scan_status (polls until complete)
You: "Show me the results — which models mentioned my brand?"Claude: → lbm_list_results
You: "What did GPT-5 say exactly?"Claude: → lbm_get_transcript
You: "Who are my competitors according to AI models?"Claude: → lbm_list_competitors

Token-Efficient Responses

All list tools return compact CSV by default instead of verbose JSON. This reduces token usage by 80–96%, keeping responses within context limits.

# Default (CSV) — 3-6 key columnscompetitor,mentions,visibility_pctCompetitor A,178,72.4Competitor B,105,42.7
# Full JSON — pass include_all_fields: true{"data": [{"competitor_id": "...", "competitor_name": "Competitor A", ...}]}

All list tools support offset and limit for pagination.

Configuration

VariableRequiredDefaultDescription
LBM_API_KEYYes—API key from llmbrandmonitor.com
LBM_API_BASE_URLNohttps://llmbrandmonitor.com/api/v1API base URL
LOG_LEVELNoinfoerror, warn, info, debug

Error Handling

Errors include actionable hints for the LLM:

ErrorHint
INSUFFICIENT_CREDITSCheck balance with lbm_get_usage. Top up at llmbrandmonitor.com/pricing
RATE_LIMITEDWait a few seconds and retry
NOT_FOUNDCall lbm_list_projects to verify the ID exists
UNAUTHORIZEDAPI key is invalid — check LBM_API_KEY

Development

bash
npm installnpm run build    # TypeScript → dist/npm test         # 126 tests

API Documentation

Full REST API docs: llmbrandmonitor.com/api-docs

License

MIT

来源：README.md，提交 12e04b6

工具

0
SourceWeft 尚未在沙箱中启动过该软件包，因此其工具未知。

版本历史

1
  1. v1.0.1最新Sep 16, 2026

评分

已满足适用检查项的 50%

该评分衡量此收录条目的完整程度，依据注册中心条目、端点与公开来源计算得出。它不是对服务器本身的评价，也不表示服务器安全或当前可用：可连通性是带日期的检查结果，评分也不包含安装量或用户评价。

  • 近 30 天内可连通25 分

    请保持端点能够响应 MCP 握手；SourceWeft 每 30 天探测一次。

  • 已声明认证方式10 分

    请在注册中心条目中声明凭据请求头，或保持端点能够响应探测，以便判定其认证方式。

  • 已列出工具15 分

    请让服务器在无凭据时响应 tools/list，或发布可在沙箱中无凭据启动的软件包。

  • 描述必需15 分
  • 图标10 分
  • 文档15 分
  • 许可证10 分