Use Session Replay over PostHog MCP

The PostHog MCP server exposes session replay to any MCP client, so your AI coding agent can find and read sessions from your editor or terminal. Investigate a bug report, check how users actually move through a feature before you change it, or confirm a fix landed – without switching to the PostHog app.

This works in any MCP client, including Cursor, Claude Code, Codex, Windsurf, and VS Code.

What you can do here

• Search recordings – find sessions by events, person properties, replay behaviors like rage clicks and dead clicks, and time range.
• Read a specific recording – pull a session's metadata by ID, including duration, interaction counts, console log counts, and a person summary. Any $session_id on an event can be used as the ID.
• Read AI summaries – list and fetch existing session summaries so the agent gets the story without watching the playback.
• Manage playlists – create collections or saved filter playlists, update them, and add or remove recordings.
• Update and delete recordings – mark recordings as viewed, or delete them one at a time or in batches of up to 100 for privacy and compliance work. Deletion is permanent, so your agent should confirm with you first.

Example prompts

Try these with your MCP-enabled agent:

• Find replays where users hit a JavaScript error on the /checkout page.
• Show me sessions from the last hour where users visited the pricing page.
• Find recordings of users on the enterprise plan who used the bulk import feature.
• Are there any replays showing users stuck in a navigation loop?
• Summarize what happened in the session for this $session_id.
• Create a collection called "Checkout friction" and add the five worst sessions to it.

Install the MCP server

The recommended way to install is with the AI wizard:

npx @posthog/wizard mcp add

The wizard supports Claude, Cursor, Windsurf, VS Code, and more. You can also configure it manually.

See the MCP server docs for full setup instructions.

Related

• Watch and share the same sessions in the web app.
• Automate against recordings and playlists with the API.
• Have AI watch recordings and turn what it sees into queryable data with Replay Vision.