Downloading Batch Export Files

Use this skill when a user wants a one-off downloadable export of PostHog data. The export is started and monitored through MCP, but the final file download uses the existing REST endpoint directly.

Available MCP tools

Do not rely on a generated MCP tool for the /download/ endpoint. That endpoint is a redirecting file download endpoint, so raw HTTP/download handling is the right interface until MCP has explicit redirect support.

Workflow

1. Choose the export shape

Ask a short clarifying question if the user did not specify the required inputs:

• model: one of events, persons, or sessions
• data_interval_start and data_interval_end: ISO 8601 datetimes; the range must be at most one week
• file.format: Parquet or JSONLines; prefer Parquet for compact analytics exports and JSONLines for line-oriented text processing
• file.compression: optional, one of zstd, gzip, brotli, lz4, or snappy. If JSONLines was chosen as format, only gzip and brotli are supported.
• file.max_size_mb: optional maximum part size in MB; set this when the user wants multiple smaller files instead of a single (potentially large) file.

For events, include and exclude are optional event-name filters. Use them only when the user asks for specific events or wants to omit specific events.

2. Start the export

Call posthog:file-download-batch-exports-create with the selected shape. The response contains an id for the export run.

Example request:

{
  "model": "events",
  "file": {
    "format": "JSONLines",
    "compression": "gzip"
  },
  "include": ["$pageview"],
  "data_interval_start": "2026-05-25T00:00:00Z",
  "data_interval_end": "2026-05-26T00:00:00Z"
}

3. Poll until completion

Call posthog:file-download-batch-exports-retrieve with the returned id.

Status handling:

When Completed, the files array contains file UUIDs. For single-file exports it usually contains one UUID. For split exports, download every UUID unless the user asked for a specific part.

4. Optionally, cancel a running export

If required by the user, a running export can be cancelled by calling posthog:file-download-batch-exports-cancel-create with the returned id.

An export that has already finished or has already failed may not be cancelled.

After cancelling an export, the id may not be used anymore and the export must start again from the beginning. However, you may still use the id to retrieve the export status (which will always be Cancelled).

5. Download files through REST

Use a direct authenticated HTTP request to the existing endpoint:

GET /api/projects/{project_id}/file_download_batch_exports/{run_id}/download/{part}/

part can be either:

• a file UUID from the files array returned by file-download-batch-exports-retrieve
• a zero-based file index, ordered by key

If there is only one file, this also works without part:

GET /api/projects/{project_id}/file_download_batch_exports/{run_id}/download/

Let the HTTP client follow the redirect, or inspect the Location header if you need the temporary signed URL. Use the same PostHog authentication context as other API calls.

6. Save, do not print, file contents

Treat the result as a file download, not a chat response. Parquet is binary and must be written as bytes. JSONLines may still be large; save it to a file rather than pasting the contents unless the user explicitly asks for a tiny sample.

Use a filename that includes the model, run ID, and part identifier when possible, for example:

posthog-events-<run_id>-<part>.jsonl.gz
posthog-persons-<run_id>-<part>.parquet

Watch-outs

• The maximum export interval is one week. Split longer user requests into separate export runs or ask which week to export.
• A run can briefly report Running after completion while file records are being created. Poll again instead of failing immediately.
• Download URLs are temporary. If a URL expires, call the REST download endpoint again for a fresh redirect.
• Do not send the signed URL to unrelated services unless the user explicitly asks; it grants temporary access to the exported file.
• If the user wants all parts of a split export, iterate over every UUID in files; do not assume part 0 is enough.
• Large batch exports may take a few minutes or even longer to complete. Suggest to the user that they can speed-up their download by including only certain events or narrowing the date range.