Intel

AIKIDO-2026-424949

headroom-ai is vulnerable to Information Disclosure

Information Disclosure Pre-CVE
Found by Aikido Intel before public disclosure or CVE publication.
Published Yesterday

64

Medium Risk

This Affects:

PYTHONheadroom-ai
0.2.2 - 0.22.4
Fixed in 0.23.0
Are you affected? Scan for Free

TL;DR

The proxy keeps a single in-process CCR context tracker that is shared across every session, project, and client. Because tracked compressions carry no workspace identity, proactive context expansion matches and injects one project's compressed content into an unrelated project's request based only on lexical keyword overlap. This exposes source and context from one workspace inside another workspace's model session. The fix adds a required per-workspace key to every tracked entry and fails closed when the workspace cannot be resolved so expansion never crosses project boundaries.

Who does this affect?

You are affected if you are using a version that falls within the vulnerable range and you run the proxy with CCR proactive context expansion enabled in default non-cache mode.

Background info

headroom-ai is vulnerable to Information Disclosure in versions 0.2.2 - 0.22.4.

How to fix this

Upgrade the headroom-ai library to the patch version.