Ghidra CVE-2026-115395

Exploitation requires the victim to actively import an attacker-supplied Ghidra theme ZIP archive through the theme import feature; the malicious archive must contain a very large number of files in its image/ subdirectory, which readZipTheme() extracts without count or size limits into ~/.config/ghidra/Ghidra*/images. Limiting factors: the attack is local and gated on user interaction (UI:P) - the target must be socially engineered into importing an untrusted theme - so there is no remote or unauthenticated trigger; it affects Ghidra versions 10.3 up to but not including 12.2, and version 12.2 enforces a 3000-entry icon cap that blocks this. Severity depends on the host filesystem's available inode pool and free disk space.

This is a genuine but low-urgency availability issue, not a high-impact priority. The provided CVSS 4.0 vector (AV:L/AC:L/AT:N/PR:N/UI:P) scores 6.8 (Medium) and is internally consistent with the description: exploitation is local, requires passive user interaction (the victim must import a theme archive), needs no privileges, and produces only an availability impact (VA:H) with zero confidentiality or integrity impact. There is NO CISA KEV listing and NO EPSS score in the supplied data, so there is no signal of active or probable exploitation - consistent with the vuln.today 'this_month' decision label rather than act_now/emergency. No proof-of-concept or exploit-code reference appears among the provided references (only NVD, the fixing commit, the 12.2 release, the GHSA advisory, and the VulnCheck advisory), so exploitation status is 'no public exploit identified at time of analysis.' The signals do not conflict: a Medium-severity, user-interaction-gated, local DoS against a developer/analyst tool. The one caveat worth verifying is reach - the GHSA notes inode/disk exhaustion can degrade OTHER system services beyond Ghidra, so impact is host-wide rather than process-local, which is why VA:H (not L) is justified.

An attacker shares a malicious Ghidra theme ZIP (e.g., via a forum, chat, or file-sharing channel) and persuades a reverse-engineer to import it. On import, readZipTheme() extracts the archive's thousands of tiny image files into ~/.config/ghidra/Ghidra*/images, exhausting the filesystem's inode pool and free disk space, which then blocks Ghidra and other local services that need to write to disk. No public exploit code was identified, but the attack requires only a crafted archive and basic ZIP-crafting skill.