1. Basic Information
- Article Title: CISA Known Exploited Vulnerabilities Catalog: CVE-2026-115731
- Publisher: CISA
- Publication Date: 2026-11-17
- Collection Date: 2026-11-18T08:00:28+09:00
- Source: CISA KEV
- Related Sources: Perforce Trust Center, Puppet releases, Red Hat, CISA Official KEV Data, Perforce CNA Record, Puppet Core 8.22.2 Release Notes
- Related CVE: CVE-2026-115731
- Severity (Author Priority): Critical
- Basis for Severity: Added to CISA KEV following confirmed active exploitation. Clients with CA-signed agent certificates can execute arbitrary code with Puppet Server service privileges. Permissive autosign may allow unauthenticated attackers to obtain certificates.
2. Quick Summary
The report handling process in Puppet Server deserializes crafted rich-data types without safe restrictions, allowing arbitrary code execution via PUT requests from clients with agent certificates.
3. Attack Flow
The preconditions described by the vendor are outlined below. Certificate acquisition paths A and B are alternatives, and it is unclear which was used in actual attacks. Both involve authentication after obtaining a certificate.
Attack Path A: Existing Certificate
- The attacker obtains a valid CA-signed agent certificate and the corresponding private key.
- The attacker sends a PUT request with a report containing crafted rich-data, aiming to achieve code execution with the privileges of the Puppet Server service account via the vulnerable report handling process.
Attack Path B: Permissive Autosign
- The attacker obtains a valid agent certificate corresponding to their own private key from an autosign feature lacking strict validation.
- After authenticating with that certificate, the attacker proceeds with the report PUT and deserialization to achieve code execution, similar to Path A.
4. Attacker Position and Execution Location
- A client that can reach the Puppet Server and authenticate using a valid agent certificate and the corresponding private key. Permissive autosign may allow attackers without prior credentials to obtain certificates, but it does not eliminate the need for authentication when performing the report PUT itself.
5. What Victims and Administrators May Observe
- Victims: May appear as configuration anomalies or impacts on credentials on managed nodes.
- Administrators: Investigation points include report PUT requests, certificate signing,
puppetserverchild processes, and modifications to the CA and manifests.
6. Success and Failure Conditions
- Success Conditions: Running an affected version, reaching the report endpoint, possessing a CA-signed certificate, and having a report containing malicious rich-data accepted.
- Failure Conditions and Mitigations: Update to a patched version and restart the server, tighten autosign settings, and restrict network access to the report endpoint.
7. What Happens upon Success
- Successful exploitation can allow code execution with Puppet Server service privileges, access to the CA and configuration data, and further compromise of managed nodes.
8. Observable Logs
- Email: None specified.
- Proxy / SWG / DNS: Unknown downloads or external outbound traffic (egress) from the Puppet Server.
- Endpoint / EDR: Execution of shells, utilities, and network clients from
puppetserver, as well as modifications to CA and manifest files. - Identity / IdP: Certificate requests and signing events, and unusual increases in agent IDs.
- SaaS / Cloud: Administrator changes or secret access in the Puppet management console.
- Network: Unusual PUT requests to
/puppet/v3/report, unknown agent certificates, and suspicious commands from the server to managed nodes.
9. Attack Success Determination
- Public Information: Inclusion in CISA KEV confirms active exploitation, but certificate acquisition paths, victim counts, threat actors, payloads, and success stages per device are not specified.
- Internal Determination Criteria Common to Paths A and B: Acceptance of a report alone does not confirm successful code execution. Verify evidence supporting unauthorized code execution with the privileges of the Puppet Server service account. CA tampering and impact on managed nodes must be checked separately.
10. Investigation Playbook
- Trigger: Unknown certificates, abnormal report PUT requests, and
puppetserverchild processes. - Initial Checks: Verify versions, autosign settings, certificate inventories, initial reports, and processes.
- Endpoint: Preserve server process trees, CA, report payloads, manifests, and file and network artifacts.
- Identity and Cloud: Review certificate issuance and revocation history, administrator sessions, and secret usage.
- Follow-On: Track command execution on managed nodes, credential access, and configuration tampering.
- Containment: Isolate endpoints, apply patches and restart, revoke certificates, rotate credentials, and re-evaluate nodes.
- Classification: Separate the stages of certificate acquisition, report submission, deserialization, code execution, and node impact.
11. Defense and Detection Ideas
- Single Event: Shell execution from
puppetserver, unexpected child processes, and signing events for unfamiliar agent certificates. - Time Series: Correlate the sequence of certificate requests -> signing -> report PUT -> child process execution -> manifest modification.
- Hunting: Investigate affected versions, permissive autosign configurations, and externally reachable report endpoints.
- Logging Gaps: KEV inclusion indicates active exploitation, but since no public IOCs exist, investigate based on behavior.
- Priority Controls: Prioritize updating to Puppet Enterprise 2026.0.1, 2025.12.1, 2023.8.13, or Puppet Core 9.1.3, 8.22.2, and restart Puppet Server. Although official Core 8 lists show a discrepancy with 8.22.1, individual 8.22.2 release notes explicitly state this CVE is fixed.
12. Facts / Inference / Hypothesis
Facts
- Perforce CNA published this as CVSS v4.0 9.4 (Critical), which is an official rating distinct from the author's priority.
- CISA added it to KEV on November 17, 2026, with a December 1, 2026 remediation deadline for covered US federal agencies.
- Red Hat rates it as Critical due to potential impact on the Puppet central CA and downstream managed systems.
Inference
- A Puppet Server compromise can extend beyond a single server to managed systems through their trust in its configuration.
Hypothesis
No additional hypotheses. Unverified items are listed under "Unknowns and Further Investigation".
13. MITRE ATT&CK Mapping
Since active exploitation payloads, interpreter usage, external exposure paths, and root certificate installation have not been publicly disclosed, MITRE ATT&CK mapping is withheld. T1553.004 is not assigned based solely on potential CA impact.
14. Unknowns and Further Investigation
- Victim counts, threat actors, IOCs, autosign exploitation rates, and the scope of impact on managed nodes have not been publicly disclosed.
- There is a discrepancy between the Core 8 official list and individual release notes mentioned above. November 3 is the publication date of the Perforce CNA record and advisory. KEV inclusion was not present as of November 3, and active exploitation was confirmed by KEV listing on November 17. We have not confirmed patched versions for Puppet Enterprise Advanced from the vendor's public documentation.
15. Impact on SOCs and Organizations
Organizations using Puppet for configuration management should not only patch servers but also review autosign settings, certificate lifecycles, report endpoint exposure, and suspicious modifications to managed nodes in an integrated manner.
16. Summary by Target Audience
- For SOCs: Correlate and monitor certificates, report PUT requests,
puppetserverprocesses, and configuration changes. - For Administrators: Apply patches, restart, tighten autosign settings, and restrict network access.
- For Users: Follow instructions if administrators reissue configurations or update credentials.
