What Permissions a Mass Deletion Tool Really Needs
When people compare deletion tools, price and speed come first. Permissions get an automatic yes. That order is backwards.
A price difference is a few tens of dollars. A permission difference can be your direct messages, your follower graph and read access to your full history. You can swap tools whenever you like. A grant you already handed over does not disappear on its own.
Three tiers of authorization
| Tier | What the tool receives | Typical implementation | Worst case |
|---|---|---|---|
| On-device parsing | Only the archive file you upload | Processing a ZIP in the browser or a local app | Data never leaves your machine |
| Scoped API access | Read your public posts and perform deletions | OAuth with narrowly requested scopes | Public content and delete rights, no direct messages |
| Full account access | Public and private content, DMs, social graph, settings | Asks for your password or very broad scopes | Equivalent to handing over a long-lived copy of your account |
The gap between the second and third tier is not technical difficulty. It is whether the tool chose restraint or convenience. A third-party tool asking for your password has no reason to exist: the platform's scope mechanism is sufficient for deleting posts.
What each tier can actually see
Tier one is capped at the archive contents, meaning whatever you chose to hand over. An archive does not carry direct message bodies, a complete follower snapshot or account settings. Even if the tool itself is untrustworthy, the exposure is limited to the file you supplied.
Tier two reads public content. A tool can walk your posts, analyze word frequency and find the ones containing phone numbers. It cannot read DMs and cannot change your account settings. The grant can be revoked from the platform's connected-apps list at any time, and the token dies immediately.
Tier three is far blurrier. Broad scopes tend to include reading DMs, viewing follows and followers, and reading the account email. None of that helps with deletion, and all of it is raw material for profiling and outreach. To judge whether a grant is excessive, count how many requested scopes have nothing to do with removing posts.
Five minutes to spot an over-broad grant
- Does it ask for your password? Anything that does is out, no exceptions.
- Count the scopes on the consent screen. Three or four is usually reasonable. Past seven or eight, read each one.
- Look for a read-DMs entry. Deleting public posts does not require message access, so its presence settles the question.
- Confirm a revocation path exists. It is only manageable if you can find the revoke button in the connected-apps list.
- Check whether the privacy policy states a retention period. A page that says it values your privacy without naming a number of days has not answered.
Step three is the most efficient single test. A tool that only deletes public posts has no functional need for message scope, so requesting it signals convenience for someone else. The full audit process is in auditing third-party app access.
Two things to do after granting access
Whichever tier you pick, go back to the connected-apps list once the deletion job finishes and revoke anything you no longer need. Grant it again next time you use it. The cost is one click.
The second item is checking logged-in devices. Revoking an app does not touch existing sessions. Both checks are covered in reviewing login devices.
About Digital Footprint Health
Digital Footprint Health (digital-footprint-health.shop) sits in tier one. You upload your X archive and everything, parsing and scoring alike, happens on your own machine with no upload and no account permission requested. Run a free check from the homepage to see your risk list; authorization only comes up if you choose to delete, with scope and pricing on the pricing page and further tool comparisons throughout the blog.
Frequently Asked Questions
Does an on-device tool still need an internet connection?
Parsing and scoring do not. They run entirely on your machine. A network request only becomes necessary when you decide to delete posts, because that action has to go through the platform's API. Checking and deleting are separate, so you can start with checking alone.
Can a tool still act on my account after I revoke access?
No. The token dies the moment you revoke, and any later request from that tool is rejected. Actions already completed are not rolled back, and there is no way to un-read data it already accessed, which is why revoking promptly matters.
Are free tools more likely to over-ask for permissions?
Not necessarily, and it is still worth watching. A free tool needs to cover its costs somehow, either through deletion fees or through the data itself. The test does not change: count how many requested scopes are unrelated to deletion. Being free is not a direct factor.
Check your own X/Twitter footprint
Free on-device scan. Your archive never leaves your computer.
Start Free CheckRelated Reads
How Tweet Deletion Tools Price Themselves: Per-Tweet, Subscription or One-Off
Tweet deletion tools look wildly different in price, but the real difference is the billing model. Per-tweet, subscription and one-off pricing produce effective unit costs that diverge sharply depending on how much you are cleaning. Here is a model comparison table, a formula for effective unit cost, and six things to confirm before you pay.
Testing a Deletion Tool Before You Pay: Trials, Credits and Refunds
Most deletion tools offer a free preview or a small paid trial, and each format verifies a different subset of claims. Here are four trial types, what they actually prove, three common refund restrictions, and a 30-minute pre-purchase check you can run yourself.
TweetDelete vs digital-footprint-health: Which One Actually Gets It?
TweetDelete is one of the most popular tweet deletion tools. But does it actually solve your problem — or just delete tweets and move on?