Connect your AI agent to your Leadmetrics account
Claude Code and Claude Desktop can read your rankings, AI-visibility reports, blog content, leads, and deliverables, and publish work once you approve it. No integration code, no credentials exposed to a browser.
Set up in under a minute
Point your client at the Leadmetrics MCP endpoint and authorize once in the browser.
$ claude mcp add --transport http leadmetrics-mcp \
https://mcp.leadmetrics.ai/mcp
# then, inside Claude Code
$ /mcpAuthorize in the browser tab that opens. No key to copy.
What your agent can reach
Four groups of tools, matched to your account permissions.
Connect and plan
Resolve the tenant you're working with and check its marketing strategy before any other call.
get_tenantFetch the full tenant profile and metadata.
strategy_get_currentGet the current marketing strategy and its recent change log.
strategy_search_versionsSearch strategy version history by range, text, or date.
strategy_get_versionGet a single strategy version by its ID.
Read
Read live tenant data across rankings, visibility, leads, and reports.
rank_keyword_listList tracked keywords and live SERP positions.
ai_search_list_tracked_promptsCheck AI-visibility prompt coverage and ranking.
lead_searchSearch leads by source, stage, date range, and text.
+ 5 more — blog_search_posts, report_search, channel_status_list, deliverable_search, activity_search
Create
Draft new records — they flow into the same review process as the rest of Leadmetrics.
blog_create_postCreate a new blog post draft.
social_create_groupDraft a social post group across platforms.
seo_keyword_createAdd an SEO keyword to the tenant dataset.
+ 2 more — deliverable_create, activity_create
Act
Update status, reassign work, and publish approved records.
activity_update_statusMark an activity as started, done, or skipped.
activity_update_assigneeReassign a task to a different user.
blog_publish_postMove a blog into the publishing flow.
+ 1 more — report_publish
MCP follows the same tenant permissions and review flow as the rest of Leadmetrics — it does not bypass the approval gate.
A few things people ask their agent to do
“Pull this month's AI-visibility report and draft fixes for the pages losing citations.”
“Find leads from the last 7 days and summarize them by source.”
“Draft and schedule next week's social posts from our content strategy.”
Security built in, not bolted on
Every MCP request runs through the same access controls and data guarantees as the rest of Leadmetrics.
OAuth 2.1 + PKCE
Browser-based authorization. No long-lived key to copy or leak.
Matches your account access
Every session carries the same permissions and data boundaries as your Leadmetrics login. An agent never reaches data you couldn't already see yourself.
DNS-rebinding protected
The connection is validated against allowed hosts and origins before any request reaches a tool.
Server-to-server only
Requests go straight from your MCP client to the server over HTTPS — nothing is exposed to browser CORS.
MCP access is enabled per account by the Leadmetrics team — contact support to turn it on.
Questions
Model Context Protocol is an open standard for connecting AI agents to external tools and data. It's what lets Claude Code or Claude Desktop call into your Leadmetrics account directly, instead of you copying data back and forth.