LLM Brand Monitor MCP Server
OfficialClick on "Deploy Server".
Wait a few minutes for the server to deploy. Once ready, it will show a "Started" state.
In the chat, type
@followed by the MCP server name and your instructions, e.g., "@LLM Brand Monitor MCP Server List all my brand monitoring projects."
That's it! The server will respond to your query, and you can continue using it as needed.
Here is a step-by-step guide with screenshots.
LLM Brand Monitor MCP Server
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:
{
"mcpServers": {
"lbm": {
"command": "npx",
"args": ["-y", "@serpstat/llm-brand-monitor-mcp"],
"env": {
"LBM_API_KEY": "lbm_your_key_here"
}
}
}
}Claude Code
claude mcp add lbm-mcp -e LBM_API_KEY=lbm_your_key_here -- npx -y @serpstat/llm-brand-monitor-mcpFrom Source
git clone https://github.com/SerpstatGlobal/llm-brand-monitor-mcp.git
cd llm-brand-monitor-mcp
npm install && npm run buildThen add to your MCP client config:
{
"mcpServers": {
"lbm": {
"command": "node",
"args": ["/path/to/llm-brand-monitor-mcp/dist/index.js"],
"env": {
"LBM_API_KEY": "lbm_your_key_here"
}
}
}
}MCP Inspector
LBM_API_KEY=lbm_your_key_here npx @modelcontextprotocol/inspector node dist/index.jsRelated MCP server: ai-visibility-mcp
Getting an API Key
Sign up at llmbrandmonitor.com
Open your Profile page
Copy your API key (starts with
lbm_)
Tools
17 tools across 4 categories:
Projects (7 tools)
Tool | What it does |
| List all brand monitoring projects |
| Get project details with prompts and models |
| Create a new project |
| Update project name, models, or monitoring settings |
| Archive a project |
| Add monitoring prompts to a project |
| Remove a prompt from a project |
Scans (3 tools)
Tool | What it does |
| Start a scan — sends prompts to LLMs and collects responses |
| Check scan progress |
| View scan history |
Results (5 tools)
Tool | What it does |
| Browse monitoring results (brand mentions, status) |
| Read the full LLM response for a specific result |
| See which competitor brands LLMs mention |
| See which URLs and domains LLMs cite |
| Competitor mention trends over time |
Models & Usage (2 tools)
Tool | What it does |
| List 350+ available LLM models |
| Check 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_competitorsToken-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 columns
competitor,mentions,visibility_pct
Competitor A,178,72.4
Competitor 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
Variable | Required | Default | Description |
| Yes | — | API key from llmbrandmonitor.com |
| No |
| API base URL |
| No |
|
|
Error Handling
Errors include actionable hints for the LLM:
Error | Hint |
| Check balance with |
| Wait a few seconds and retry |
| Call |
| API key is invalid — check |
Development
npm install
npm run build # TypeScript → dist/
npm test # 126 testsAPI Documentation
Full REST API docs: llmbrandmonitor.com/api-docs
License
MIT
Available Tools
17 toolslbm_add_promptsA
WHEN TO USE: To add one or more monitoring prompts to a project. Prompts are the questions asked to LLMs (e.g. "What are the best tools for X?"). Max 50 per request, 100 per project total. REQUIRES: project_id from lbm_list_projects. RETURNS: Array of created prompts with prompt_id. NEXT STEP: Call lbm_run_scan to run a scan with the new prompts.
Inputs: prompts (required), project_id (required).
lbm_archive_projectADestructiveIdempotent
WHEN TO USE: To archive a project that is no longer needed. Archived projects are hidden but not deleted. REQUIRES: project_id from lbm_list_projects. Confirm with user before archiving.
Inputs: project_id (required).
lbm_create_projectA
WHEN TO USE: To create a new brand monitoring project. RETURNS: Created project object with generated id. NEXT STEP: Call lbm_run_scan to start monitoring.
Inputs: domain (required), models (required), prompts (required), language (required), +4 more.
lbm_delete_promptADestructiveIdempotent
WHEN TO USE: To remove a specific prompt from a project. REQUIRES: project_id and prompt_id. Get prompt_id from lbm_get_project (prompts array). NOTE: Project must have at least 2 prompts; cannot delete from archived project. Confirm with user before deleting.
Inputs: prompt_id (required), project_id (required).
lbm_get_historyARead-onlyIdempotent
WHEN TO USE: To see competitor mention trends over time — which competitors were mentioned by LLMs across multiple scans and how their visibility changed. Use for competitive trend analysis and reporting. REQUIRES: project_id from lbm_list_projects. RETURNS: Historical competitor mention statistics per date, with per-model breakdowns. NOTE: Response structure may differ between time_range and start_date/end_date params.
Inputs: project_id (required), limit, offset, end_date, +4 more.
lbm_get_projectARead-onlyIdempotent
WHEN TO USE: To get full details of a specific brand monitoring project including all prompts. REQUIRES: project_id from lbm_list_projects. RETURNS: Project object with id, name, description, status, models, prompts array, schedule, and stats.
Inputs: project_id (required).
lbm_get_scan_statusARead-onlyIdempotent
WHEN TO USE: To check the status of a running or completed scan. REQUIRES: project_id and scan_id from lbm_run_scan or lbm_list_scans. RETURNS: Scan object with status (pending/running/completed/failed), progress, start/end times, and result count. NEXT STEP: When status="completed", call lbm_list_results to see monitoring results.
Inputs: scan_id (required), project_id (required).
lbm_get_transcriptARead-onlyIdempotent
WHEN TO USE: To read the full verbatim LLM response for a specific monitoring result. Use when the user wants to see exactly what an LLM said about their brand. REQUIRES: project_id and result_id from lbm_list_results. RETURNS: Full transcript text, model name, prompt text, and metadata (brand_mentioned, sentiment, links found).
Inputs: result_id (required), project_id (required).
lbm_get_usageARead-onlyIdempotent
WHEN TO USE: To check credit balance and usage statistics. Call this before lbm_run_scan to confirm the user has enough credits. RETURNS: Current credit balance, credits used this period, credits per scan estimate, and usage breakdown by model and project.
Inputs: end_date, start_date.
lbm_list_competitorsARead-onlyIdempotent
WHEN TO USE: To see which competitor brands were mentioned by LLMs in responses to this project's prompts. Shows competitive landscape as seen by AI models. REQUIRES: project_id from lbm_list_projects. RETURNS: Compact CSV with competitor name, mention count, frequency % (default limit: 20). Set include_all_fields=true for full JSON. Pass higher limit only if user explicitly asks for more.
Inputs: project_id (required), tags, limit, offset, +1 more.
lbm_list_linksARead-onlyIdempotent
WHEN TO USE: To see which URLs and domains LLMs cited in their responses — useful for understanding what sources AI models trust for this topic. REQUIRES: project_id from lbm_list_projects. RETURNS: Compact CSV with domain, mentions, unique_urls (default limit: 20). Set include_all_fields=true for full JSON with individual URLs. Pass higher limit only if user explicitly asks for more.
Inputs: project_id (required), tags, limit, domain, +3 more.
lbm_list_modelsARead-onlyIdempotent
WHEN TO USE: To discover which LLM models are available for brand monitoring scans. Call this before creating a project or running a scan if the user wants to choose specific models. RETURNS: Compact CSV with model_id, name, provider, web_search (default). Set include_all_fields=true for full JSON with pricing.
Inputs: limit, offset, include_all_fields.
lbm_list_projectsARead-onlyIdempotent
WHEN TO USE: To get a list of all brand monitoring projects for the current user. Use as the FIRST step to discover available projects before working with scans or results. RETURNS: Compact CSV with id, brand_name, status, visibility_pct, models_count, prompts_count (default). Set include_all_fields=true for full JSON. NEXT STEP: Use lbm_get_project with a specific project id for full details.
Inputs: limit, offset, status, include_all_fields.
lbm_list_resultsARead-onlyIdempotent
WHEN TO USE: To get monitoring results for a project — how LLMs responded to brand monitoring prompts. REQUIRES: project_id from lbm_list_projects. RETURNS: Compact CSV with result_id, prompt, model, brand_mentioned, status (default limit: 20). Set include_all_fields=true for full JSON. NEXT STEP: Use lbm_get_transcript with a result_id to read the full LLM response text.
Inputs: project_id (required), tags, limit, offset, +3 more.
lbm_list_scansARead-onlyIdempotent
WHEN TO USE: To see the scan history for a project — all past and current scans with their statuses. REQUIRES: project_id from lbm_list_projects. RETURNS: Compact CSV with scan_id, status, created_at, results_count (default). Set include_all_fields=true for full JSON.
Inputs: project_id (required), limit, offset, status, +1 more.
lbm_run_scanADestructive
WHEN TO USE: To start a new monitoring scan for a project — sends prompts to all configured LLMs and collects responses. REQUIRES: project_id from lbm_list_projects. Project must have at least one prompt (use lbm_add_prompts if needed). RETURNS: Created scan object with scan_id and status="pending". NEXT STEP: Poll lbm_get_scan_status with scan_id until status="completed". CAUTION: This SPENDS user credits. Always confirm with the user before running.
Inputs: project_id (required), models, prompt_ids, use_web_search.
lbm_update_projectAIdempotent
WHEN TO USE: To update project name, models, or auto monitoring settings. Only provided fields are updated. NOTE: brand_name and prompts cannot be modified via this endpoint. Use lbm_add_prompts/lbm_delete_prompt for prompt changes. REQUIRES: project_id from lbm_list_projects. RETURNS: Updated project object.
Inputs: project_id (required), models, project_name, auto_monitoring.
Tool Schema Changelog
Recent tool additions, removals, and schema changes observed during successful MCP inspections.
Loading tool schema changelog…
TDQS
Scored across 17 tools
Each tool has a clearly distinct purpose, targeting specific resources (projects, prompts, scans, results, competitors, links, history, usage) with unique actions (list, get, create, update, archive, add, delete, run). There is no functional overlap.
All tools follow a consistent `lbm_verb_noun` pattern in snake_case. Verbs are standard and nouns are appropriately singular or plural, making it easy to infer tool behavior from names.
With 17 tools, the scope is well-balanced. It covers full lifecycle management (CRUD for projects, prompts, scans, results) and adds analytics (competitors, links, history, usage) without unnecessary bloat.
The surface is nearly complete, covering all major operations. Minor gaps exist: no dedicated tool to update prompts (only add/delete) and no permanent project deletion (only archive). These are workable but not critical.
Maintenance
Related MCP Connectors
AI-visibility monitoring for your brand across ChatGPT, Claude, Perplexity & Gemini.
Brand visibility auditing across LLMs, AI search, and answer engines with GEO reports and scores.
Measure brand mentions in ChatGPT, Gemini, Claude and Perplexity answers and diagnose GEO.
Track brand visibility in ChatGPT, Gemini and Google AI answers: competitors, cited sources, checks.
Related MCP Servers
- AlicenseAqualityDmaintenanceEnables brand visibility monitoring across major AI platforms like ChatGPT, Claude, Gemini, and Perplexity. It allows users to track visibility scores, analyze competitor data, and receive actionable insights to improve AI-generated brand recommendations.1611 npm1MIT
- AlicenseAqualityDmaintenanceTrack brand visibility across ChatGPT, Perplexity, Claude, and Gemini.611MIT
- AlicenseAqualityDmaintenanceTrack how your brand appears in AI-generated answers across ChatGPT, Perplexity, and other AI models. Analyze visibility, sentiment, citations, and domain rankings with 31 tools — including analytics reports, chat inspection, query analysis, and full CRUD for brands, prompts, tags, and topics.1797 npm2MIT
- AlicenseBqualityCmaintenanceEnables AI agents to check brand mentions across AI search surfaces like ChatGPT, Claude, Gemini, Perplexity, and Google AI Overviews using natural language queries.41MIT