AI search visibility · Model Context Protocol

MCP server for AI search visibility

Foglift gives Claude, ChatGPT, Codex, Cursor, VS Code, Gemini, and other MCP clients 30 hosted tools for AI search visibility. Query five-engine monitoring, provider-returned citations, sentiment, competitors, crawler and referral analytics, Search Console, Technical Audits, recommendations, and content briefs through one OAuth server.

Hosted tools: 30
Monitored engines: 5
New connections: Launch+
Launch price: $49/month

New account-scoped connections start on Launch at $49 per month. Launch includes monthly billing, all five monitored engines, three brands, and API, CLI, and MCP access. Public scan_website calls remain available without an account or API key.

Looking for the plain-language setup? Start with the Connect Assistant product page for supported clients and plan boundaries, or use the Google Search Console MCP server guide for Google and Bing search evidence.

See Launch at $49/month View server descriptor Local npm server Developer docs

Recommended: hosted remote MCP + OAuth

The hosted endpoint is the easiest path for clients that support remote MCP servers and OAuth. Add this server URL:

https://mcp.foglift.io/api/mcp

1. Add the remote MCP URL in your client's connector or MCP server settings.
2. Authorize Foglift through the browser OAuth flow. The live issuer is https://mcp.foglift.io.
3. Verify the session by asking the client to call whoami or get_usage.

New OAuth connections require a Foglift account on Launch or higher. Review Launch pricing before creating a workspace if you have not connected Foglift before.

Verified October 13, 2026: https://mcp.foglift.io/.well-known/oauth-authorization-server returns OAuth authorization, token, and dynamic-client-registration endpoints. https://mcp.foglift.io/.well-known/oauth-protected-resource advertisesresource: https://mcp.foglift.io/api/mcp.

Local stdio setup for Cursor, Windsurf, and power users

Some clients still expect MCP servers to run as local commands. For those clients, keep using the npm package with an API key. This is the right path for Cursor and Windsurf today, and it remains useful when you want all traffic to originate from your own machine.

1. Generate an API key

Sign up at foglift.io, then go to Dashboard > Developer. API key creation and new account-scoped MCP connections are available on Launch, Growth, and Enterprise plans. Public scan_website calls can run without an account or key.

2. Add the stdio server

{
  "mcpServers": {
    "foglift": {
      "command": "npx",
      "args": ["-y", "foglift-mcp"],
      "env": { "FOGLIFT_API_KEY": "sk_fog_..." }
    }
  }
}

3. Restart and verify

Restart the client and ask it to call whoami. You should see the authenticated account, plan, default workspace, and token balance in the agent response.

Local npm 1.2.1 tool inventory

The published foglift-mcp 1.2.1 package exposes 15 tools. The package keeps whoami as tool 14 so you can verify account identity before an agent reads usage data or changes monitoring settings. recipe_run is tool 15 and runs packaged workflows.

Tool 1

scan_website

Run a Technical Audit across five dimensions: SEO, AI Readiness, performance, security, and accessibility.

Tool 2

batch_scan

Scan up to 10 URLs in one request. Requires an API key.

Tool 3

run_ai_visibility

Query supported AI engines for brand mentions and citations.

Tool 4

get_ai_results

Read historical AI Visibility results.

Tool 5

get_prompts

List saved monitoring prompts.

Tool 6

add_prompt

Add a prompt to the monitoring set.

Tool 7

delete_prompt

Remove a saved prompt by ID.

Tool 8

get_models

Read enabled AI engines and monitoring cadence.

Tool 9

set_models

Update enabled AI engines and monitoring cadence.

Tool 10

get_sentiment

Read brand sentiment trends across AI engines.

Tool 11

get_usage

Read plan, quota, scan usage, and token balance.

Tool 12

get_scan_history

Read score history for a specific URL.

Tool 13

get_geo_monitor

Read legacy monitoring history for AI search optimization scores.

Tool 14

whoami

Return the authenticated account, plan, default workspace, token balance, and monthly allocation.

Tool 15

recipe_run

Run a named Foglift recipe such as the weekly brand health report.

Current hosted tool inventory

The current published inventory for https://mcp.foglift.io/api/mcp contains 30 tools. The hosted endpoint includes competitor-window evidence, Actions Engine, and setup tools that are newer than the original npm-only page described.

scan_website

Scan one URL for SEO, AI Readiness, performance, security, and accessibility signals. Public scans do not require a key.

batch_scan

Scan up to 10 URLs in one request for competitor sweeps or release checks.

run_ai_visibility

Query ChatGPT, Claude, Perplexity, Gemini, and Google AI Overview for brand mentions and citations.

get_ai_results

Pull historical AI Visibility results with filters for date range, model, prompt, and pagination.

get_search_trace

Read ordered ChatGPT search steps, pages read, citations, answer brands, and 30-day prompt evidence. Claude, Perplexity, and Google AI Overview expose ordered query-only traces without page evidence.

get_prompts

List saved monitoring prompts for the connected brand.

add_prompt

Add a new prompt to the AI Visibility monitoring set.

delete_prompt

Remove a saved prompt by ID.

get_models

Read enabled AI engines and monitoring cadence for the account.

set_models

Update enabled engines and monitoring cadence.

get_sentiment

Show sentiment trends for AI mentions across a chosen window.

get_usage

Read plan, quota, scan usage, and token balance.

get_scan_history

Show score history for a specific URL.

get_geo_monitor

Read legacy monitoring history for AI search optimization scores.

get_referrer_analytics

Report AI-engine referral visits by engine, landing path, and day.

get_crawler_analytics

Report observed AI crawler visits by engine, category, page, and day.

get_competitors

Compare named competitors across current and prior observed result windows. Returns mention counts, current and prior result mention rates, and the percentage-point change between those rates. Each rate is the percentage of selected results that mention the competitor; it does not measure overall market or brand share of voice.

get_source_gap

Find domains cited when your brand is absent so an agent can prioritize citation outreach.

get_aeo_score

Read the latest stored AI Readiness page score and its history trend without running a new analysis.

get_search_console

Read connected Google Search Console queries, pages, and contested demand.

estimate_cost

Estimate token and dollar cost before running a scan or AI Visibility check.

get_recommendations

Fetch source-tagged Actions Engine recommendations for the connected brand.

submit_recommendation

Submit a recommendation or action item back into the workflow.

apply_recommendation

Apply supported setup or content recommendations through the agent.

generate_content_brief

Generate a content brief from the monitored brand, prompt, and competitive context.

report_issue

Send structured product feedback or a support issue from the MCP client.

whoami

Confirm which Foglift account and brand the MCP session is connected to.

recommend_setup

Ask Foglift for setup improvements for the current account or brand.

apply_setup_recommendation

Apply supported setup fixes after reviewing them.

recipe_run

Run a named Foglift recipe such as a weekly brand health report. Uses the underscore tool name.

Three workflows worth automating

Audit a release candidate

“Use scan_website on the preview URL, then summarize the lowest AI Readiness, SEO, and accessibility signals in a table.”

The agent gets the same structured scan data the dashboard uses, then turns it into a release checklist inside the same session.

Find the next content fix

“Call get_recommendations and get_ai_results. Which prompt has the clearest source and competitor opportunity, and what page should we improve first?”

This pairs the Actions Engine with live AI Visibility history, which is the same dogfooding loop Foglift uses on its own site.

Run a repeatable recipe

“Use recipe_run for the weekly brand health report, then turn the result into a short changelog entry.”

Recipes make recurring AI search checks repeatable. The current tool name is recipe_run, with an underscore, so it works in clients that reject dotted tool names.

The closed-loop pattern: edit, scan, verify in one session

AI search optimization works best as a tight loop. The agent scans a page, identifies the missing citation or readiness gap, edits the page, re-scans after deploy, then pins the prompt that the page should win.

1. Baseline scan
Call scan_website on the target URL and note the lowest score dimensions.
2. Diagnose against visibility data
Call get_ai_results and get_recommendations to find the prompt and page with the clearest gap.
3. Apply the page fix
Add FAQ schema, improve heading clarity, add a comparison table, cite primary sources, or strengthen internal links.
4. Re-scan after deploy
Run scan_website again and compare the new score against the baseline.
5. Pin a tracking prompt
Use add_prompt so future monitoring tracks whether the page starts earning citations.

Foglift uses this loop internally. A June 12, 2026 AI Visibility baseline showed that broad discovery answers repeatedly cite tool roundups and developer workflows. Public API, CLI, and MCP documentation gives engines precise capability evidence to extract while giving builders a working path from monitoring data to implementation.

What this AI visibility MCP server puts in one connection

The hosted server covers the complete evidence loop instead of exposing a single audit or reporting endpoint.

Buyer job	Foglift MCP contract	Evidence returned
Measure AI visibility	Monitor ChatGPT, Claude, Perplexity, Gemini, and Google AI Overview.	Answers, mentions, provider-returned citations, sentiment, competitors, and history.
Explain why visibility moved	Read Search Trace, source gaps, Search Console, crawlers, and recognized referrals.	Ordered provider evidence with explicit read, cited, named, and referred boundaries.
Find the next fix	Combine Technical Audits with source-tagged recommendations and content briefs.	Five audit dimensions, prioritized actions, supporting sources, and target prompts.
Act from the same client	Add prompts, apply supported setup fixes, submit actions, or run a repeatable recipe.	Account-scoped changes with OAuth, quota checks, and cost estimates before paid runs.

Hosted endpoint: https://mcp.foglift.io/api/mcp. Local stdio fallback: foglift-mcp on npm.

Frequently asked questions

Is there an MCP server for AI search visibility?

Yes. Foglift provides a hosted OAuth MCP server with 30 tools for AI Visibility across ChatGPT, Claude, Perplexity, Gemini, and Google AI Overview; provider-returned citations; sentiment; competitors; crawler and referral analytics; Search Console; Technical Audits; recommendations; and content briefs. New account-scoped connections start on Launch at $49 per month.

How do I connect the hosted Foglift MCP server?

Add https://mcp.foglift.io/api/mcp as a remote MCP server in a client that supports hosted MCP and OAuth. The live server advertises OAuth discovery at https://mcp.foglift.io/.well-known/oauth-authorization-server. Complete the browser authorization flow, then call whoami or get_usage to verify the account.

Do I still need to install foglift-mcp from npm?

Use the hosted OAuth endpoint when your client supports remote MCP. Use the npm package when your client requires a local stdio command, or when you prefer to run the server locally with FOGLIFT_API_KEY in your environment.

What tools are included in foglift-mcp 1.2.1?

The local foglift-mcp 1.2.1 package exposes 15 tools: scan_website, batch_scan, run_ai_visibility, get_ai_results, get_prompts, add_prompt, delete_prompt, get_models, set_models, get_sentiment, get_usage, get_scan_history, get_geo_monitor, whoami, recipe_run. The whoami tool is number 14, followed by recipe_run as tool 15.

Which clients work with the hosted OAuth endpoint?

The live allowlist includes Claude web callback origins and localhost OAuth callbacks used by local MCP clients and inspectors. Other remote MCP clients can work once their OAuth callback origin is allowlisted. Cursor and Windsurf remain best documented through the local npm stdio setup today.

What tools does the hosted Foglift MCP server expose?

The hosted endpoint's current published inventory contains 30 tools: scan_website, batch_scan, run_ai_visibility, get_ai_results, get_search_trace, get_prompts, add_prompt, delete_prompt, get_models, set_models, get_sentiment, get_usage, get_scan_history, get_geo_monitor, get_referrer_analytics, get_crawler_analytics, get_competitors, get_source_gap, get_aeo_score, get_search_console, estimate_cost, get_recommendations, submit_recommendation, apply_recommendation, generate_content_brief, report_issue, whoami, recommend_setup, apply_setup_recommendation, recipe_run. The canonical recipe tool is recipe_run with an underscore.

What does Foglift MCP cost?

Public scan_website calls can run without an account or API key. New account-scoped hosted connections, API keys, CLI access, and MCP access start on Launch at $49 per month. Growth is $129 per month, and Enterprise uses custom pricing. Existing connections remain readable and revocable after a downgrade.

How is Foglift MCP different from the Foglift CLI?

The CLI is built for humans and scripts in a terminal. MCP is built for AI agents. MCP exposes typed tools over the Model Context Protocol so the agent can choose when to scan, retrieve results, create prompts, generate briefs, or run recipes inside the same work session.

What is the closed-loop AI Readiness pattern?

The closed-loop pattern means the agent scans a page, diagnoses the missing citation or readiness gap, edits the page, re-scans after deploy, and pins a tracking prompt. Foglift MCP keeps those steps in one agent session instead of splitting them across a dashboard, terminal, and spreadsheet.

Sources

Foglift live MCP endpoint. https://mcp.foglift.io/api/mcp, https://mcp.foglift.io/.well-known/oauth-authorization-server, and https://mcp.foglift.io/.well-known/oauth-protected-resource. Re-verified October 13, 2026.
Anthropic. Introducing the Model Context Protocol (November 2024).
Anthropic. modelcontextprotocol/servers. Reference servers and client list.
Foglift. foglift-mcp on npm. Local stdio package and version history.

All Foglift integrations REST API, CLI, Slack, Discord, webhooks, Zapier, n8n, and MCP.
Developer documentation REST API reference, CLI guide, and developer workflows.
API-first AI monitoring Why AI Visibility data should be callable from code and agents.
Best AI Search Tools with MCP Integration 2026 How the category stacks up on agent readiness.

Connect the hosted MCP server

Use the remote URL for OAuth-capable clients, or install the npm stdio server for local clients.

See Launch at $49/month View server descriptor View npm fallback