Back to MCP Servers

Peec AI MCP Server

Official MCP server for Peec AI. Query brand visibility, sentiment, citations, and share of voice across ChatGPT, Claude, Gemini, Perplexity, and more.

Analytics by Peec AI OAuth2 active
Overview

Peec AI is an AI search analytics platform that tracks how brands appear across ChatGPT, Claude, Gemini, Perplexity, Google AI Overviews, Google AI Mode, Microsoft Copilot, and Grok. The official Peec AI MCP server exposes the same data the Peec dashboard uses, letting AI assistants query visibility metrics, sentiment, citations, source content, and competitor positioning, plus mutate project configuration (prompts, topics, tags, brands, classifications) directly from chat.

The server is hosted by Peec at https://api.peec.ai/mcp over Streamable HTTP. It supports both OAuth 2.0 (browser consent flow) and Personal Access Tokens for headless setups. The current tool set covers read operations (reports, chats, sources, actions) and write operations (create, update, delete, bulk) for project configuration. Destructive operations are flagged and write tools require organization-owner access.

Notable features include built-in workflow prompts (weekly pulse, engine scorecard, topic heatmap, campaign tracker), opportunity-scored Actions recommendations, and access to the scraped markdown content of source URLs cited by AI engines. The MCP is available on all paid Peec plans at no additional cost.

Tools

Tool Description
list_projects Lists all Peec projects the authenticated user can access.
list_brands Lists tracked brands (own and competitors) with domains and aliases.
list_topics Lists topic groupings used to organize prompts.
list_tags Lists cross-cutting labels attached to prompts.
list_models Lists the AI engines configured for a project.
list_model_channels Lists stable engine channel identifiers.
list_prompts Lists prompts, filterable by topic or tag.
list_chats Lists individual AI responses over a date range.
get_chat Retrieves full chat content including brands mentioned, sources, and sub-queries.
list_search_queries Lists search sub-queries issued by AI engines.
list_shopping_queries Lists product/shopping sub-queries and extracted products.
get_url_content Returns scraped markdown content of indexed source URLs.
list_domain_classifications Lists custom domain classification definitions.
list_url_classifications Lists custom URL classification definitions.
get_brand_report Returns visibility, sentiment, position, and share of voice metrics for brands.
get_domain_report Returns source domain retrieval and citation metrics.
get_url_report Returns URL-level retrieval and citation metrics.
get_actions Returns opportunity-scored action recommendations from Peec Actions.
get_project_profile Reads brand profile data used for prompt generation.
create_brand Creates a tracked brand with domains and aliases.
update_brand Updates brand details (name, domains, aliases, regex, color).
delete_brand Soft-deletes a brand. Destructive.
create_brands / delete_brands Bulk create or delete up to 50 brands per call.
create_prompt Creates a new prompt for a country code.
update_prompt Updates a prompt's topic and tags.
delete_prompt Soft-deletes a prompt; cascades to chats. Destructive.
create_prompts / delete_prompts Bulk create or delete up to 50 prompts per call.
create_tag / update_tag / delete_tag Create, rename, or soft-delete a tag.
create_tags / delete_tags Bulk tag operations for up to 50 tags.
create_topic / update_topic / delete_topic Create, rename, or soft-delete a topic grouping.
create_topics / delete_topics Bulk topic operations for up to 50 topics.
create_domain_classification / delete_domain_classification Define or permanently delete a custom domain classification.
assign_domain_classification / unassign_domain_classification Assign or clear a domain classification override.
create_url_classification / delete_url_classification Define or permanently delete a custom URL classification.
assign_url_classification / unassign_url_classification Assign or clear a URL classification override.
set_project_profile Updates brand description, industry, positioning, and audience data used for prompt generation.
Setup Guide

Prerequisites

  • An active Peec AI account on any paid plan with at least one project
  • A compatible MCP client (Claude Desktop/Web, Claude Code, Cursor, VS Code with GitHub Copilot, Windsurf)

Claude Desktop / Claude Web

Open Settings, find the Peec AI connector, and click Connect. You will be redirected to Peec AI to sign in via OAuth. Note that only workspace admins can add connectors in Claude.

Claude Code (CLI)

OAuth (recommended):

claude mcp add peec-ai --transport http https://api.peec.ai/mcp

Personal Access Token (headless):

claude mcp add peec-ai --transport http https://api.peec.ai/mcp \
  --header "Authorization: Bearer YOUR_PEEC_PAT"

Cursor

Go to Cursor Settings, Tools & Integrations, MCP. Add a custom MCP server:

  • URL: https://api.peec.ai/mcp
  • Transport: Streamable HTTP

For PAT auth, add an Authorization: Bearer YOUR_PEEC_PAT header.

VS Code (GitHub Copilot)

Create or edit .vscode/mcp.json:

{
  "servers": {
    "peec-ai": {
      "type": "http",
      "url": "https://api.peec.ai/mcp"
    }
  }
}

With PAT:

{
  "servers": {
    "peec-ai": {
      "type": "http",
      "url": "https://api.peec.ai/mcp",
      "headers": {
        "Authorization": "Bearer YOUR_PEEC_PAT"
      }
    }
  }
}

Windsurf

Open Windsurf Settings, MCP, click Add Server, and enter https://api.peec.ai/mcp. Authenticate via Peec AI on first use.

Generating a Personal Access Token

In app.peec.ai, go to API Keys, Personal Access Tokens, and generate a new token. Attach it as a Bearer token in the Authorization header.

Verifying

Ask your assistant: "List my Peec AI projects". If a list returns, the connection is working.

Use Cases
  • Pull a weekly brand visibility pulse across ChatGPT, Claude, Gemini, and Perplexity, comparing share of voice against named competitors.
  • Investigate a sentiment drop by drilling from get_brand_report into individual list_chats and get_chat calls to read the exact AI responses driving it.
  • Audit which source domains and URLs AI engines cite about your category, then read the scraped markdown via get_url_content to spot content gaps.
  • Programmatically expand prompt coverage by bulk-creating prompts, topics, and tags for new geos or product lines.
  • Surface opportunity-scored next actions via get_actions and feed them into Slack, email, or a spreadsheet for the marketing team.
Example Prompts
  • "List my Peec AI projects, then show me the brand visibility report for the last 14 days grouped by engine."
  • "What happened to our share of voice on ChatGPT last week, and which prompts drove the biggest changes?"
  • "Pull the top 20 source domains cited about our category on Perplexity and fetch the markdown for the top 3 URLs."
  • "Create 10 new prompts in topic 'Pricing comparisons' tagged 'US-2026' for country code US."
  • "Show me the latest Peec Actions sorted by opportunity score and summarize the top five recommendations."
Pros
  • Official, provider-hosted server: no self-hosting, no third-party fork to trust.
  • Comprehensive coverage with both read (reports, chats, sources, actions) and write (CRUD on brands, prompts, tags, topics, classifications) tools, including bulk operations.
  • Supports both OAuth 2.0 and Personal Access Tokens, with destructive operations flagged for safety.
  • Included on all paid Peec plans at no extra cost; works out of the box with Claude, Cursor, VS Code, and Windsurf.
Limitations
  • Requires an active paid Peec AI subscription; not usable without an account and project data.
  • Write/configuration tools require organization-owner access, limiting what non-admin teammates can change.
  • Specific input schemas are not exhaustively documented in the public docs; some parameter details must be discovered at runtime.
Alternatives
  • Profound for AI search visibility analytics (separate platform; no first-party MCP server documented).
  • AthenaHQ and similar AI brand monitoring tools; most do not yet ship official MCP servers.
  • Community-built thein-art/mcp-server-peecai on GitHub, which wraps the same Peec API for self-hosted setups but is not maintained by Peec.