Privacy Analysis With Your X Archive: How to Safely Review Your Own Tweets

Your X data archive is the single source of truth for any privacy check. Before you can clean up, you have to understand what the archive holds and why the analysis must stay on your own machine.

What the archive actually contains

Most people think of tweets as text. The archive shows the rest: timestamps, location tags, linked devices, ad topics, and media. As the breakdown of tweets.js explains, the post file is only one layer. Ad data and media metadata carry their own signals, and those are easy to miss if you only skim the text.

Why on-device analysis matters

Privacy tools face a hard conflict: they handle your most sensitive data, yet many upload it to servers. A trustworthy check should never need to upload your archive, call X's read API for more data, or store your tweets in a database. With on-device parsing, the archive is read locally and only touched by X's write API when you choose to delete.

How to review safely

1. Request the archive from your X settings; it arrives as a ZIP.
2. Open it with a tool that runs locally. Pick the format that fits your workflow; CSV versus JSON explains the trade-off.
3. Let it scan for phone numbers, addresses, locations, and sensitive topics across tweets, ads, and media.
4. Read the risk list and decide what to delete. Nothing leaves your computer during this step.
5. When you delete, the tool talks to X's write API on your behalf, and only then.

The signals people miss

Tweets get the attention, but two quieter files matter. Ad-interest data shows what X thinks you like, which can reveal health, politics, or finances you never posted about directly. Media metadata can carry device and location clues from the photos you attached. A good check looks at all of it, not just the visible text.

How to read the risk list

When the check finishes, you get a list sorted by risk. Do not panic at the length; read the labels first, then decide what to delete.

High (red): phone number, address, email, live location. These must go, and you should file a search removal request too, or the snapshot stays up for months.

Medium (yellow): sensitive topics, controversial stances, identity-revealing interest tags. Not all need deleting, but if you are job hunting or switching teams, clear the most glaring ones.

Low (green): daily chatter, reposts, memes. Unless you care a lot, leave them. Spend your effort on red and yellow for the best cleanup per minute.

One practical tip: sort by time, newest first. Recent exposures are most likely still in search caches and most likely to be seen by new colleagues or friends. Old posts matter too, but they get dug up far less often.

The list is not there to scare you; it is there to queue the work. Once risks are tiered, you know which few items ten minutes can fix today. Many people delay cleanup because a long list feels like nowhere to start; with red, yellow, and green in place, the first step is obvious. Start with red, and you will often clear the bulk of real danger in a single sitting, then move down at a relaxed pace. Progress beats perfection, and a shorter risk list tomorrow is the only measure that matters.

Get started

The archive is the raw material; on-device analysis is the safe way to turn it into a clear risk picture. Start a free footprint check that never uploads your data, and review your own tweets without handing them to anyone.