On November 17, the US Cybersecurity and Infrastructure Security Agency (CISA) added a critical Puppet Enterprise vulnerability to its Known Exploited Vulnerabilities (KEV) catalog. The flaw is tracked as CVE-2026-115731. The CVE record and fixed versions were published on November 3, so the KEV listing came exactly two weeks later. That is a short gap between a patch shipping and CISA saying the bug is being exploited in the wild, and the software involved makes it worse. Puppet Server is the system many organizations use to push configuration to every machine they manage.
CISA set a remediation due date of December 1, 2026. It has not said who is exploiting the flaw or how. The catalog lists known ransomware use as "Unknown."
What Puppet Does, and Why This Bug Matters
Puppet is a configuration management platform. A central Puppet Server tells a fleet of managed machines, called agents, how they should be configured. Each agent identifies itself to the server with a certificate signed by Puppet's own certificate authority (CA). After an agent applies its configuration, it sends a report back to the server describing what happened.
The vulnerability is in how the server handles those reports. The Debian Security Tracker describes it as a deserialization flaw in report processing. Deserialization means turning data received over the network back into objects the program can use, and here the data was trusted when it should not have been. The result is that a client holding any CA-signed agent certificate can execute code on Puppet Server. CISA's catalog entry lists two weakness types: CWE-502 (deserialization of untrusted data) and CWE-94 (code injection).
In practice, the bar for exploitation is any enrolled machine. Compromising one ordinary agent anywhere in the fleet is enough to reach the server. The CVSS 4.0 vector from Puppet, the CVE Numbering Authority (CNA) that published the record, reflects this: network attack vector, low attack complexity and low privileges required. The score is 9.4, Critical.
The Autosign Problem
The most important caveat in the advisories concerns certificate autosigning. Puppet's CA can be set to approve agent certificate requests automatically. That is convenient when new machines are being spun up constantly, but it is dangerous when the requests are not strongly verified.
According to the advisory text in the NVD record and the Debian tracker, if the CA autosigns requests without strong verification, an unauthenticated attacker on the network can get a valid agent certificate and then use it to exploit the flaw. Red Hat says the same in its own assessment: networks with permissive automatic certificate signing may let unauthorized remote actors obtain credentials. On those deployments, the "privileges required" in the CVSS score can be obtained by anyone who can reach the CA.
What an Attacker Gets
Red Hat rates the issue Critical. It says successful exploitation gives command execution as the Puppet service account, and that this could compromise the central CA and every system Puppet manages. Puppet Server is the machine that tells every other machine what to run, so control of it can quickly become control of the whole fleet.
Red Hat also notes that Red Hat Satellite processes client reports by default. Satellite customers should therefore consider the vulnerable code path reachable unless Red Hat's guidance says otherwise.
Affected and Fixed Versions
According to the NVD record, the affected Puppet Enterprise releases are:
- 2023.8.0 up to, but not including, 2023.8.13
- 2025.0.0 up to, but not including, 2025.12.1
- 2026.0.0 up to, but not including, 2026.0.1
The NVD record points to a vendor advisory on Perforce's support portal. CISA's catalog lists the vendor as Perforce.
The open-source OpenVox fork of Puppet also shipped fixes. As of November 11, SmartScope reported fixes in OpenVox Server 9.2.0 and 8.18.0 and OpenVox Agent 9.3.0 and 8.32.0. The same report says Japan's JPCERT/CC urged users to update in its weekly report of November 11. It also points to a second, separate OpenVox advisory, GHSA-q4hm-33rc-572g, which covers client impersonation. Admins applying the fixes should keep that one in mind as well.
Debian has released updates through DLA-4884-1 (the fixed puppet-agent package for bookworm) and DSA-6608-1 (for trixie), both referenced on its security tracker.
What We Don't Know
CISA adds a vulnerability to KEV when it has reliable evidence of exploitation. The catalog entry itself does not identify attackers, targets or methods, and CISA has published no such details for this CVE. None of the sources reviewed here attributes the activity to any group or country. The ransomware field is "Unknown," which means CISA has not confirmed ransomware use. It does not mean ransomware has been ruled out.
What It Means for Defenders
Who is affected: anyone running Puppet Enterprise in the version ranges above, OpenVox Server or Agent below the fixed releases, distribution-packaged Puppet on Debian, or Red Hat products such as Satellite that process Puppet client reports. Federal agencies bound by CISA's directive have until December 1. CISA's required action is to apply the vendor mitigations under BOD 26-04 or stop using the product. Everyone else should treat that date as an upper limit, not a target.
Patch first. Move Puppet Enterprise to 2023.8.13, 2025.12.1 or 2026.0.1, depending on your release line. Move OpenVox Server to 9.2.0 or 8.18.0 and OpenVox Agent to 9.3.0 or 8.32.0. Debian users should apply DLA-4884-1 or DSA-6608-1 as appropriate. Red Hat and Satellite customers should follow Red Hat's guidance for CVE-2026-115731.
Audit autosigning. If your CA automatically approves certificate requests without strong verification, this flaw can be exploited without credentials. Check your autosign configuration. Accept automatic approval only where requests are actually verified, and treat any blanket or pattern-only autosign rule as an emergency.
Review who holds certificates. Any machine with a CA-signed agent certificate could reach the vulnerable code path. Look for certificates you don't recognize, retired hosts whose certificates were never revoked, and signing activity that doesn't match your provisioning records.
Assume the worst if you were exposed. Red Hat warns that exploitation could compromise the central CA and every managed system. If you ran a vulnerable server with permissive autosigning, or you suspect an agent was compromised, the patch does not close the incident. Review the Puppet server for unauthorized changes made under the service account. Check whether the configuration code and data the server distributes to agents have been altered. Consider rebuilding trust in the Puppet CA: reissue the CA and agent certificates and rotate any secrets the server can reach.
Limit network reach. Puppet Server and its CA should be reachable only from the networks where managed agents actually live, never from the open internet.
Federal contractors working toward CMMC should note that a configuration management server can typically reach every system in a CUI enclave. A compromise of that server is a compromise of the whole enclave, so this patch belongs at the front of the queue.