Analytics

PostHog with Claude & ChatGPT—PostHog MCP server

Product events and persons on the unified contact timeline, identity-resolved with your CRM, support, and email, read-only. Works in Claude, ChatGPT, Cursor, and any MCP-compatible agent.

What it looks like

What does PostHog data look like in Claude?

Connect PostHog once. Your agent reads product events and usage on the unified profile—identity-resolved with your CRM, support, email, and website events—through one MCP server.

MCP

PostHog, exposed as MCP tools.

Claude · pathbound MCP connected
you
Which trial users hit our activation event this week that no one has followed up with?
list_events { event: "activation", since: "-7d", identified_only: true } 23 contacts · 460ms
posthog 23 fired activation · 18 in trial
identified all 23 resolved to a contact
list_contacts { lifecycle: "trial", source: "posthog" } 18 contacts · 410ms
cross-ref 9 with no email reply in 14d
gmail last touch > 2 weeks ago
claude
9 trial users fired activation this week with no follow-up in two weeks—including Jane Reyes (acme.com) and Tom Liu (bridge.io). Activation came from PostHog; the silence is from Gmail; the trial stage is from HubSpot—all on one profile.
signal Activated and silent is the highest-converting nudge window. Reach out today.
claude_desktop_config.json
{
  "mcpServers": {
    "pathbound": {
      "url": "https://mcp.pathbound.ai/mcp"
    }
  }
}
PostHog tools out of the box · read
get_contact unified profile incl. PostHog person properties
get_contact_timeline product events alongside CRM, email, and support
list_events filter PostHog events by type, date, identified_only
aggregate_data activation and usage breakdowns—no raw event dumps
list_contacts filter by lifecycle, source, or recent activity
create_contact_note log findings on the unified contact

…plus 20 more—26 MCP tools total, across contacts, companies, events, and conversations.

Setup

How do I connect PostHog to Claude?

Paste a project API key. Drop a snippet into your MCP client. Ask a question.

  1. 01

    Sign up for Pathbound

    Create an account at pathbound.ai. Free during public alpha.

  2. 02

    Connect PostHog

    In the dashboard, go to Integrations → PostHog → Connect. Paste a PostHog Personal API key and your project. Sync pulls persons and product events and resolves them to the unified contact by distinct_id and email.

  3. 03

    Add Pathbound to your MCP client

    Drop the snippet below into your Claude Desktop config (or in Claude.ai web: Settings → Connected Apps → Add custom MCP → enter https://mcp.pathbound.ai/mcp). The same URL works for ChatGPT, Cursor, Cline, and any MCP client that supports OAuth.

    claude_desktop_config.json
    {
      "mcpServers": {
        "pathbound": {
          "url": "https://mcp.pathbound.ai/mcp"
        }
      }
    }
  4. 04

    Run a test prompt

    Ask the agent: “How many product events did [email protected] fire this week?” The agent calls list_events via MCP and returns the count.

Why Pathbound

PostHog tells you what happened. Pathbound tells you to whom—across everything.

A direct PostHog connector shows your agent product events. Pathbound shows your agent the customer—PostHog events stitched together with your CRM, support, email, and website events—through one unified profile served over the Model Context Protocol.

Product behavior in context

Feature usage and activation events land on the same timeline as deal stage, support tickets, and emails. Your agent sees what they did in the product next to everything else—not in a separate analytics tab.

Identity-resolved

A PostHog person (distinct_id, email) resolves to the same unified contact you have in your CRM, Gmail, and website events. A direct connector only knows PostHog’s own IDs.

Dashboards stay in PostHog

PostHog powers your product team’s dashboards; Pathbound puts the same events on the contact profile your AI agent (and your AE or CSM) reads before the next action. Different consumers, same source.

Built for LLMs

Token-efficient response shapes and server-side aggregation mean agents get usage summaries and counts, not a raw event firehose that blows the context window.

FAQ

Common questions.

Is this an official PostHog MCP server?

Pathbound is an independent platform. We connect to PostHog via its API using a key you generate, and expose your data through one Model Context Protocol server alongside your other sources. Not affiliated with PostHog.

Does Pathbound write to PostHog?

No. The PostHog integration is read-only—it pulls persons and product events into the unified profile and never writes anything back to PostHog.

How is this different from PostHog’s own dashboards or API?

PostHog is product analytics for your team’s dashboards and queries. Pathbound puts the same events on the per-contact profile your AI agent, AE, or CSM reads before the next action—identity-resolved with your CRM, email, and support. Different consumers, same source.

How are PostHog events matched to a contact?

By identity resolution on distinct_id and email. A person’s product behavior lands on the same unified profile as their CRM record, conversations, and web activity.

Which MCP clients does this work with?

Claude.ai, Claude Desktop, Claude Code, ChatGPT (with custom connectors), Cursor, Cline, Continue—any client that supports the Model Context Protocol with OAuth.

Is my customer data secure?

Tenant-isolated and OAuth-scoped per user, with encryption at rest via our infrastructure provider’s default configuration. GDPR-aligned; SOC 2 readiness in progress. Pathbound does not sell customer data, and we don’t call AI providers on your behalf—your MCP client does, under your account and the AI provider’s terms.

What about HubSpot, Intercom, Gmail?

Same MCP server, same tools—just connect a different source. See the integrations directory.

Try it on your own data

Within five minutes, your agent is querying real customer context.

Sign up, connect one CRM or drop the events snippet, and point your MCP client at mcp.pathbound.ai/mcp.