Loading vulnerability…
CVE-2026-31852 affects the jellyfin/jellyfin-ios repository's GitHub Actions workflow, specifically code-quality.yml. The workflow is vulnerable to arbitrary code execution when processing pull requests from forked repositories. Because the workflow executes in a context with nearly all write permissions, untrusted pull request-controlled content can be used to run attacker-supplied code within the GitHub Actions environment. This is a CI/CD workflow misconfiguration rather than an application code flaw. The described consequence is repository-level compromise of jellyfin/jellyfin-ios, exposure of privileged secrets available to the workflow, and abuse of cross-repository token access within the Jellyfin GitHub organization.
Mallory correlates every CVE against your assets, your vendors, and active adversary campaigns. Know which vulnerabilities matter for you, not just which ones are loud.
What it means. What to do now. Patch path, mitigations, and the assume-compromise checklist.
What an attacker gets, and what they’ve been doing with it.
If you can’t patch tonight, do this now.
Patch, then assume compromise.
1 valid exploit after Mallory filtered fakes, detection scripts, and README-only repos.
This 100-file repository is an automated research/lab snapshot of Jellyfin for iOS, not a conventional standalone exploit. It primarily contains an Expo/React Native iOS media-client application (JavaScript/TypeScript), bundled WebView bridge scripts, download and native media-player components, Jest tests, iOS Objective-C++ project files, and Fastlane/Ruby release automation. The README explicitly describes it as a disposable artifact for reproducing published GitHub Actions workflow vulnerabilities with dummy secrets. The security-relevant artifact is .github/workflows/code-quality.yml. Its eslint job is triggered by pull_request_target, explicitly checks out github.event.pull_request.head.sha, then invokes npm ci and npm run lint. This combines a privileged base-repository workflow context with untrusted pull-request content. A malicious PR can alter package metadata/dependency lifecycle behavior (or project scripts) so code runs during npm ci or the lint step. The job does not explicitly map secrets into its environment, but pull_request_target commonly supplies a base-context GITHUB_TOKEN whose effective permissions depend on repository configuration; thus impact can include token abuse and repository/API actions, in addition to runner command execution. The separate automation.yml also uses pull_request_target but does not check out or execute PR code in the displayed workflow. No malicious command, callback host, reverse shell, exfiltration logic, fixed target server, or exploit payload is present. Accordingly, this is best characterized as a proof-of-concept vulnerable workflow/reproduction artifact rather than weaponized exploit code. The actual Jellyfin application accepts user-configured server URLs, loads them in a WebView, injects a NativeShell compatibility bridge, handles media playback, and downloads files from configured Jellyfin instances; these normal application capabilities are distinct from the GitHub Actions workflow issue.
Products and vendors Mallory has correlated with this vulnerability. Open in Mallory to drill down to specific CPE configurations and version ranges.
Vendor-confirmed product mapping. Mallory continuously reconciles this list against your asset inventory.
10 sources tracked across advisories and community write-ups. News coverage will land here when it surfaces.
No news coverage yet. Advisories and community discussion only.
Query your assets running an affected version, and investigate the blast radius.
Every observed campaign linking this CVE to a named adversary.
Malware families riding this exploit, with evidence and IOCs.
YARA, Sigma, Snort, and vendor rules, auto-deployed to your SIEM.
Cross-references every affected SKU, including bundled OEM variants.
Community discussion across Reddit, Mastodon, and other social sources.