All blog posts
Research 3 min read

What the Commonveil Security Review Actually Found

The recorded security review identified 24 issues and prompted permission, identity and test changes, while leaving independent cryptographic assurance as separate work.

By Comutability Published Update from
Illustration of security findings being examined through a magnifying glass, fixed and checked with regression tests.

Research record from 12 November 2026. This article discusses published project research and implementation evidence, not independent certification.

The recorded security review identified 24 issues and prompted permission, identity and test changes, while leaving independent cryptographic assurance as separate work.

Why a list of findings is valuable

Security confidence improves when a project records problems rather than hiding them. Commonveil's November review includes an outside report and a numbered findings list, CV-01 through CV-24. The repository says fixes were prepared on a named security branch with regression tests intended to fail against the previous behaviour.

That is an important engineering milestone, but it needs precise wording. It is not a statement that all installations were upgraded, that every change has received a second independent review, or that the newer Secure Net cryptography has been certified. A branch, a fix and a production deployment are distinct pieces of evidence.

The highest-risk themes

Some reported issues concerned identity ownership and account takeover. A signed-in session could replace an owner key without the right additional proof. Disconnected-device tokens could still be used for certain sign-in changes. The first registered member on a new self-hosted site could accidentally receive administrator status. The response introduced stronger owner-key checks, security epochs for sessions and a one-time bootstrap code for the first administrator.

Other findings concerned encrypted community access. Removing a device or revoking channel access needed to invalidate related secrets. Polling and other read paths had to enforce the same permission checks as normal message retrieval. These are examples of boundary conditions that can be missed when tests follow only the most common UI path.

Operational failures also count

The review includes medium-severity findings about invitations, moderation scope, search visibility, scheduled check-ins, dependency locking and update status. A community service can expose data or mislead users without a headline cryptographic failure. Race conditions during invitation use or an updating indicator that never clears are part of trustworthiness.

The design response therefore reaches into Docker builds, permission calculations and automated regressions, not just an encryption library.

How testing changed

The documentation describes regression tests that reproduce the old failures, then verify the intended fix. It also records a later blank screen caused by JavaScript declaration ordering in an experimental protocol release. That failure had escaped handler tests because the actual signed-in interface was not rendered during the test run. A static order check was then added, and the fix was confirmed in a browser against a local site.

This illustrates why successful tests need several layers: unit tests, real backend integration, browser journeys, deployment checks and human review. Mock-heavy suites should not be treated as complete proof of production behaviour.

What remains unproven

Security review is a continuing process. The sources explicitly distinguish the review findings from a qualified independent cryptographic audit of the latest protocol, and the repository continues to document known limitations. The next steps are to verify which fixes reached each deployment, repeat regression tests after migration and commission deeper expert evaluation for sensitive cryptography.

Publishing the findings gives users and contributors an honest way to judge progress. It is better to state a specific improvement with evidence than to promise absolute security.

Primary sources and further reading

Explore other Comutability research ยท Read about the organisation

Sources and further reading

Explore the evidence and original project records behind this article.