Disclaimer: This content was generated using AI. While we strive for accuracy, please verify critical information with official sources.
Sources: National Vulnerability Database, CISA KEV, Exploit-DB
CVE-2026-112780 Overview
CVE-2026-112780 is an allowlist bypass vulnerability in the Grav flat-file content management system (CMS). The flaw affects Grav versions from 2.0.15 before 2.2.2 and is tracked under CWE-749: Exposed Dangerous Method or Function. Authenticated users with page-edit permission can rename Flex form fields using the replace-name@ directive to smuggle data-*@ attributes past the callable allowlist check. The bypass permits invocation of arbitrary public static methods, including Folder::delete, Folder::move, and Utils::download. Successful exploitation allows attackers to delete directories, overwrite files on disk, or read sensitive files from the server filesystem.
Critical Impact
Authenticated page editors can achieve arbitrary file deletion, overwrite, and read on the host filesystem by abusing Flex form field renaming to invoke dangerous static methods.
Affected Products
- Grav CMS version 2.0.15 and later
- Grav CMS versions prior to 2.2.2
- Grav installations exposing Flex form editing to non-administrative users
Discovery Timeline
- 2026-10-19 - CVE-2026-112780 published to the National Vulnerability Database (NVD)
- 2026-10-22 - Last updated in NVD database
Technical Details for CVE-2026-112780
Vulnerability Analysis
Grav's Flex form framework processes form field definitions that can map user input to internal callables. The framework enforces an allowlist to block invocation of unapproved static methods. The vulnerability arises because the replace-name@ directive renames a field after the allowlist check runs. An attacker renames a benign field to a data-*@ form that resolves to a sensitive callable at execution time. The dispatcher then invokes the renamed target without re-validating it against the allowlist.
The reachable methods include Folder::delete for directory removal, Folder::move for arbitrary file overwrite through relocation, and Utils::download for reading files from paths the web user can access. Together these primitives give an authenticated editor read, write, and destructive control over the Grav installation and adjacent filesystem content.
Root Cause
The root cause is an order-of-operations defect in the Flex form field handler. The callable allowlist is evaluated against the original field name, while the replace-name@ directive mutates the effective field identifier afterward. No post-rename re-validation enforces the allowlist on the final resolved callable.
Attack Vector
Exploitation requires network access to the Grav admin interface and an account with page-edit permission. The attacker submits a crafted Flex form payload containing a replace-name@ entry and a data-*@ directive referencing a sensitive static method. The server executes the smuggled callable with the privileges of the Grav web process. Refer to the GitHub Security Advisory GHSA-9rq7-4cch-f22j and the VulnCheck Advisory for Grav 2.0.15 for payload specifics.
No verified proof-of-concept code is published. See the linked advisories
for technical write-ups and remediation detail.
Detect Identity Attacks the Moment Authentication Fails
Singularity™ Identity surfaces credential misuse, lateral movement, and privilege escalation in real time, so you can contain the breach before it spreads.
Stop Identity AttacksDetection Methods for CVE-2026-112780
Indicators of Compromise
- Unexpected replace-name@ or data-*@ tokens in Flex form submissions captured in web server or application logs.
- Unexplained deletions or relocations of files under the Grav user/ directory tree.
- Outbound responses from the admin interface delivering file contents outside normal download workflows, consistent with abuse of Utils::download.
Detection Strategies
- Inspect HTTP POST bodies to the Grav admin endpoints for form field names containing the @ suffix, particularly replace-name@ and data-*@ patterns.
- Correlate page-editor session activity with filesystem change events on the Grav host to surface unauthorized Folder::delete or Folder::move effects.
- Hunt for Grav PHP process events that read sensitive files such as .env, configuration YAML, or user credential stores outside normal request flows.
Monitoring Recommendations
- Enable verbose logging on the Grav admin plugin and forward logs to a centralized analytics platform for retention and query.
- Monitor file integrity across the Grav webroot and user/accounts/ directory to detect unauthorized writes or deletions.
- Track creation and modification of editor-role accounts, since this vulnerability requires only page-edit privileges rather than administrative rights.
How to Mitigate CVE-2026-112780
Immediate Actions Required
- Upgrade Grav to version 2.2.2 or later immediately on all production and staging instances.
- Audit all accounts holding page-edit permission and remove unnecessary privileges.
- Review web server and admin logs for replace-name@ or data-*@ artifacts dating back to the installation of version 2.0.15.
Patch Information
The maintainers resolved CVE-2026-112780 in Grav 2.2.2. The fix enforces the callable allowlist against the final resolved field name after replace-name@ processing, preventing smuggled directives from reaching dangerous static methods. See the GitHub Security Advisory GHSA-9rq7-4cch-f22j for the referenced commits and release notes.
Workarounds
- Restrict the Grav admin interface to trusted networks using a reverse proxy, VPN, or IP allowlist until the upgrade completes.
- Revoke page-edit permissions from accounts that do not require content authoring, reducing the exploitable user population.
- Place the Grav webroot on a filesystem mount with least-privilege ownership so that Folder::delete and Folder::move abuse cannot reach adjacent sensitive directories.
# Upgrade Grav to a fixed release using the bundled CLI
cd /path/to/grav
bin/gpm selfupgrade -f
bin/gpm version
# Confirm the reported Grav version is 2.2.2 or later
Disclaimer: This content was generated using AI. While we strive for accuracy, please verify critical information with official sources.
Sources: National Vulnerability Database, CISA KEV, Exploit-DB

