Penfold Desk is a shared inbox tool. Each customer gets a workspace, and threads belong to a workspace.
You are in workspace 204. The thread list pages with a cursor the server issues, and the front end sends it back untouched.
A support thread on Penfold's own helpdesk has a screenshot in it from a customer in another workspace, showing their browser address bar. The cursor in it reads eyJvIjo4ODF9.
One change to this request triggers the vulnerability. The challenge is to use your security intuition to work out what to change. Find the value the server shouldn’t trust, change it, and send it. The response tells you whether you were right.
Hint. Read threads from a workspace that is not yours.
“A token the server issued is a token the server can trust when it comes back.”
Take the workspace from the session on every request and use the cursor only for position. If the cursor must carry scope, sign it and verify it, or encrypt it so it is genuinely opaque, and re-check the caller against whatever it decodes to. Add a test that pages a list using a cursor issued to another tenant and expects a refusal.
Anywhere a value is round-tripped because it looks like the server's own: pagination cursors, continuation tokens, state parameters, encoded filter blobs, view state, resume tokens, anything base64 in a URL. Opaque is not the same as signed. The check is to decode it and see whether it contains a decision rather than a position.
One Request Away is a daily series of challenges. They’re like code review, but for HTTP requests: in each one you’re one request away from triggering a vulnerability, and your goal is to identify what you need to change.