CVE-2026-77349
PMO eval output could execute script in the review viewer
cody-hutson · pmo-platform
First documented disclosure · Review checkpoint
Summary
Model or skill outputs could break out of the generated review page.
Severity and exploitation evidence
Medium · 5.0/10 · CVSS 3.1. Severity source.
Exploitation not established: Public evidence describes a weakness, source correction or controlled demonstration; malicious exploitation against users is not established.
Technical severity is not a measure of observed attacks or of risk to your deployment. Attempts, demonstrations, and successful compromise retain the qualifications in the source.
How the weakness works
Unescaped JSON enters an inline script, while unsafe DOM render sinks remain exploitable after transport-only escaping.
Conditions required
An attacker-influenced evaluation output and an operator generating and opening the viewer.
Potential impact
Read embedded evaluation data, alter displayed results or tamper with feedback.
Affected versions and configurations
3.48–3.69.
Fix and mitigation status
3.69.1.
Practical next steps
Upgrade and redeploy the corrected evaluation skill; regenerate affected review pages.
Operating lesson
Escape for both transport and the final rendering context.
Evidence and limits
Maintainer corrected CVSS 3.1 7.1 to 5.0 and confirmed render-sink fixes were necessary. No third-party session or credential held by the viewer is established. CNA publication 2026-11-20 is distinct from the earliest reviewed public source. GitHub_M assessment: CVSS 3.1 5.
Sources
- GitHub_M CVE authority record
Primary source ·
- pmo-platform original advisory
Primary source ·