How to Build Your Own Shadowban Detector

When Reddit shadowbans an account, the user is not notified. They can still log in, vote, and submit comments. To the shadowbanned user, everything looks completely normal. To the rest of the world, their profile returns a 404 Not Found response.

The Engineering Challenge

Building a reliable shadowban detector requires polling the Reddit API from an unauthenticated, isolated proxy. If you poll the user's profile using their own OAuth token, the API will return a 200 OK because the user exists in their own localized, isolated state.

The Core Verification Loop

"Detecting shadowbans early saves agencies thousands of dollars. A ghosted account operating for a month is pure wasted labor."

Our system, SubSafe, utilizes a distributed proxy network to perform 'clean room' verification.

• Step 1: The user requests a shadowban check via the SubSafe dashboard.
• Step 2: An isolated worker node, using an IP that has never interacted with the user's account, fires a pure GET request to reddit.com/user/{username}/about.json.
• Step 3: If the JSON response returns an error: 404, and the account is known to exist, the user is definitively shadowbanned.
• Step 4: For post-level removals (ghosting), the worker checks the specific thread JSON from an unauthenticated state to verify if the comment is visible in the comment tree.