Attack ChainAIDerived
Hypothetical attack flow derived from CVE metadata – not a confirmed exploit chain
-
AccessDeliver crafted ASS + malicious Lua script
-
DeliveryVictim opens subtitle project in Aegisub
-
ExploitMetadata auto-loads referenced Automation script without trust prompt
-
ExecutionArbitrary code executes as Aegisub process
-
ImpactAttacker gains local code execution
Vulnerability Assessment
| Weakness |
CWE
OWASP Top 10:2025
A05:2025 - Injection
ASVS 5.0
V1 - Encoding and Sanitization Metacharacter neutralization
|
| Exploitation Conditions | Exploitation requires the victim to open, within Aegisub 3.2.0-3.4.2, a crafted .ass subtitle project whose 'Automation Scripts' metadata references an attacker-supplied Lua Automation script; Aegisub auto-loads that script without any trust/authorship confirmation. The attacker must get both the ASS file and the referenced script to the victim (e.g., distributed together), except on 3.4.0-3.4.2 where inconsistent handling of embedded NUL characters between extension validation and filesystem operations lets a single crafted ASS/Lua polyglot reference and execute itself, removing the need for a separate script file. Limiting factors: the vector is local (AV:L) and requires user interaction (UI:R) to open the file, it does not affect default behavior of any network service, and only versions 3.2.0 through 3.4.2 are affected (fixed in 3.5.0). |
| Impact | This is a genuine code-execution flaw but its real-world urgency is moderate, not critical. The CVSS 3.1 vector (AV:L/AC:L/PR:N/UI:R) correctly models a social-engineering file-open attack rather than a remotely reachable service bug: the attacker cannot reach a victim over the network and instead must convince the user to open a malicious ASS subtitle project (UI:R), which then auto-loads a referenced Lua Automation script with full code execution (C:H/I:H/A:H). The 7.8 'High' score is driven almost entirely by the triple-high impact, while the local vector and required user interaction meaningfully cap exploitability. No EPSS score, CISA KEV listing, or public POC is present in the supplied data, so there is no public exploit identified at time of analysis and no evidence of active or widespread exploitation - the risk is latent and depends on tricking a specific user. The vulnerability is credibly confirmed by the vendor's own GitHub Security Advisory (GHSA-67hq-5vgg-6f23) and three linked remediation commits, so the technical validity is high even though threat-intel exploitation signals are absent. Net: treat as a priority for the Aegisub user community (particularly fansub/typesetting workflows where ASS files are routinely exchanged), but it is not an internet-facing emergency. |
| Exploit ScenarioDerived | An attacker publishes or shares a seemingly legitimate .ass subtitle project (e.g., on a fansub forum or download site) bundled with a malicious Lua Automation script it references, or on 3.4.0-3.4.2 a single ASS/Lua polyglot using an embedded NUL to self-reference. When the victim opens the file in a vulnerable Aegisub (3.2.0-3.4.2), the script is loaded automatically with no trust prompt and runs arbitrary code at the privileges of the Aegisub process. No public exploit code is identified in the provided data, but the attack requires only standard file-sharing and the victim opening the file. |
Get an e-mail when this CVE is exploited, patched or re-scored. Free account.
Alert meAnalysisAIv1 – Oct 09, 2026
Arbitrary code execution in Aegisub 3.2.0 through 3.4.2 occurs when a victim opens a crafted .ass subtitle project whose 'Automation Scripts' metadata references an attacker-supplied Lua Automation script, which Aegisub auto-loads without prompting the user to trust the script or its author and runs with the privileges of the Aegisub process. Versions 3.4.0 through 3.4.2 add a single-file attack path in which inconsistent handling of embedded NUL characters between extension validation and filesystem operations lets a crafted ASS/Lua polyglot reference and execute itself, removing the need to deliver a separate script file. The flaw is rated CVSS 7.8 (AV:L/AC:L/PR:N/UI:R/S:U/C:H/I:H/A:H), making it a local, user-interaction-driven issue that does not affect any network service; there is no public exploit identified at time of analysis, so real-world urgency is moderate rather than critical, and Aegisub 3.5.0 fixes the vulnerability.
RemediationAI
Vendor-released patch: Aegisub 3.5.0, which resolves both the unconditional auto-loading of referenced Automation scripts and the inconsistent embedded-NUL handling described in GitHub advisory GHSA-67hq-5vgg-6f23 (upstream fix commits 0a3073d59f10432eeebea65018b02ef4d0a87fff, b5bf23979e7453f7b1cc8ffeb82b3cc2d2ca75e6, and e4099055711cce327840870c050ce46e58f4aee0); upgrading to 3.5.0 or later is the only complete fix. No official workaround is documented in the advisory, so until the upgrade can be deployed, apply manual compensating controls and accept their trade-offs: treat any .ass/.ssa project from an untrusted source as executable content and open it only inside a sandbox, container, or low-privilege virtual machine, since the referenced Automation script runs with the full privileges of the Aegisub process; inspect untrusted projects in a plain text editor and delete the 'Automation Scripts' line from the [Script Info] section before opening them in Aegisub, which removes the auto-load trigger but technically alters the project and also disables any legitimate scripts the project relied on; or neutralize the Automation subsystem entirely by emptying or renaming Aegisub's automation directories (for example the automation/autoload and automation/include folders under the install or user data directory), which prevents referenced scripts from being found and executed at the cost of losing all Automation functionality. Because the vector is local and requires the user to open the crafted file (AV:L with UI:R), e-mail and web gateways that block or quarantine .ass/.ssa attachments and downloads, file-extension and MIME filtering, and user awareness training all materially reduce exposure while 3.5.0 is being rolled out.
Affected VersionsENISA
Affected Packages
Source: OSV (Open Source Vulnerabilities)
Affects your stack
Affects 1 package in euvd
Checking this CVE against your own stacks and servers is part of the Basic plan and up.
Technical ContextAI · show details
Aegisub is a cross-platform advanced subtitle editor whose Automation subsystem executes Lua scripts to manipulate subtitle timing, styling, and formatting. The ASS subtitle format stores project settings in a [Script Info] section, and one accepted key, 'Automation Scripts', lists script filenames that Aegisub loads automatically when the project is opened, with no trust prompt or authorship check, which is what turns an apparently benign subtitle file into a code-execution vector. The root cause class is CWE-158 (Improper Neutralization of Null Byte or NUL Character): in Aegisub 3.4.0 through 3.4.2, the extension check applied to a referenced script and the later filesystem open handle embedded NUL bytes inconsistently, so a filename whose apparent extension validates as .lua can be truncated differently at the point of actual execution, enabling the single-file ASS/Lua polyglot. The upstream fix commit 0a3073d59f10432eeebea65018b02ef4d0a87fff changes auto filename = p.filename().string(); to std::string filename = p.filename().string().c_str(); in libaegisub/common/fs.cpp, with the comment 'Be consistent with std::filesystem functions which use C strings internally', closing the mismatch between the validated name and the name used on disk; additional fixes are carried in commits b5bf23979e7453f7b1cc8ffeb82b3cc2d2ca75e6 and e4099055711cce327840870c050ce46e58f4aee0. The CPE cpe:2.3:a:typesettingtools:aegisub:*:*:*:*:*:*:*:* confirms the vendor and product namespace, while the affected range is bounded by the advisory to 3.2.0 through 3.4.2.
Know when Aegisub gets its next vulnerability
- Watch products and get an email when something new is published
- Check this CVE against your own stack and servers
- Exploit and PoC references, threat intelligence from 22 sources
No credit card · 7-day full trial
Severity by source
AV:L/AC:L/PR:N/UI:R/S:U/C:H/I:H/A:H
Attacker needs no privileges but must deliver a file the victim opens (AV:L/UI:R); auto-loading the script is reliable (AC:L) and yields full code execution (C:H/I:H/A:H) within the same process scope (S:U).
Primary rating from Vendor (GitHub_M).
CVSS VectorVendor: GitHub_M
Lifecycle Timeline
2DescriptionCVE.org
Aegisub is a cross-platform advanced subtitle editor. From 3.2.0 to 3.4.2, Aegisub automatically loads Automation scripts referenced by Automation Scripts metadata in ASS subtitle projects without asking whether the user trusts the scripts or their authors. An attacker can distribute a crafted ASS file together with a referenced malicious Automation script, and opening the AS file executes arbitrary code with the privileges of the Aegisub process. From 3.4.0 to 3.4.2, inconsistent handling of embedded NUL characters between extension validation and filesystem operations additionally allows a crafted ASS/Lua polyglot to reference and execute itself as a single-file variant. The vulnerability is fixed in Aegisub 3.5.0.
Technical POC Analysis
Two chained defects, both mapped by the advisory to CWE-158 (Improper Neutralization of Null Byte or NUL Character) and CWE-829 (Inclusion of Functionality from Untrusted Control Sphere): Uncondition…
Full technical POC analysis - root cause, step-by-step exploitation and patch breakdown - is part of the Pro and Business plans.
Executive ViewAI
Aegisub, the subtitle-editing tool used by localization, post-production and captioning teams, will execute attacker-supplied Lua code with the full rights of the logged-in user the moment an employee opens a crafted .ass subtitle project, requiring no separate malware delivery, no network service exposure and no elevated account. Versions 3.4.0 through 3.4.2 reduce the attack to a single shared file, so one inbound subtitle asset from a freelancer, partner studio or public repository can compromise an editor's workstation and anything reachable from it, including saved credentials, mounted shares and other production systems. Because subtitle files are treated as routine, low-risk creative content, the exposure is behavioral rather than infrastructural, and leadership should act on asset handling and fleet versioning rather than perimeter defenses.
Apply vendor patch.
Threat Intelligence Overview
How this decision was reached
The engine reads each signal below and applies a fixed list of rules from top to bottom. The first rule that matches sets the recommended action – like a statement, every line is shown.
What every action level means
SSVC AssessmentENISA
Priority Score
Attack Technique
Remote Code Execution represents the critical moment when an attacker successfully runs arbitrary code on a target system without physical access.
EUVD-2026-95988
NVD
CVE.org
EPSS
GitHub
VulDB
Exploit-DB