Executive Summary
In mid-2026, the Git project disclosed six distinct security vulnerabilities affecting all widely used Git versions, addressed in Git 2.56.1 and backported to multiple maintenance releases. The flaws span arbitrary code execution via integer truncation during recursive clones (CVE-2026-111760), path traversal through malicious server-advertised HEAD symrefs (CVE-2026-113408), stack corruption from oversized GPG signature arrays (CVE-2026-114467), out-of-bounds reads from crafted binary patches (CVE-2026-114469), heap corruption in large merge operations (CVE-2026-105437), and unauthorized filesystem access through fast-import streams (CVE-2026-114466). Exploitation requires only that a developer clone or merge from a malicious or compromised repository, making this a direct supply-chain threat to any organization using Git-based workflows, CI/CD pipelines, or hosted services such as GitHub Enterprise Server.
These vulnerabilities illustrate the expanding attack surface of software supply chains, where a single interaction with a poisoned repository can compromise developer workstations, build infrastructure, and cloud-hosted code platforms. As threat actors increasingly target developer tooling to achieve broad downstream impact, organizations must treat version-controlled infrastructure as a critical security boundary.
Why This Matters Now
Software supply-chain attacks are at an all-time high, with adversaries pivoting from application exploits to developer tooling. These six Git CVEs are exploitable through ordinary clone and merge workflows, meaning any CI/CD pipeline, GitHub Actions runner, or automated build system that processes external repositories is immediately at risk until patched. Regulatory frameworks including PCI DSS 4.0 and NIST CSF now explicitly require supply-chain risk controls, increasing compliance exposure for unpatched environments.
Attack Path Analysis
An attacker crafts a malicious Git repository exploiting CVE-2026-111760 (integer truncation in index handling) to replace a submodule's .git file with a symbolic link, triggering arbitrary code execution during a recursive clone on a victim developer machine or CI/CD pipeline. From that initial foothold, the attacker leverages attacker-controlled Git hooks and configuration to escalate privileges within the build environment, pivot laterally across interconnected build systems and cloud workloads, establish persistent command-and-control channels, exfiltrate source code and secrets, and ultimately disrupt software supply chain integrity by injecting malicious artifacts into downstream pipelines.
Kill Chain Progression
This analysis maps confirmed threat intelligence to the full cloud kill chain to show where defensive gaps would emerge as an attack progresses.
Initial Compromise
Description
The attacker hosts a malicious Git repository containing a specially crafted index with an extremely long path (CVE-2026-111760) that bypasses path validation via integer truncation, replacing a submodule's .git file with a symbolic link to attacker-controlled configuration containing malicious hooks. When a developer or automated CI/CD system performs a recursive clone of this repository, Git executes attacker-controlled hook scripts. Additionally, CVE-2026-113408 allows a malicious server to advertise a HEAD symref with path-traversal components during a protocol v2 clone, enabling writes outside the intended reference directory. Cloud Native Security Fabric (CNSF) with inline enforcement and real-time inspection at the network perimeter could have detected anomalous clone traffic patterns; however, the primary control gap is the absence of repository trust validation and egress policy preventing connections to untrusted Git servers.
Related CVEs
CVE-2026-111760
CVSS 0An integer truncation in Git's index handling allows an attacker to supply an extremely long path that bypasses path validation, enabling replacement of a submodule's .git file with a symbolic link to attacker-controlled content, resulting in arbitrary code execution during recursive clones.
Affected Products:
Git SCM Git – < 2.56.1, < 2.55.1, < 2.54.1, < 2.53.1, < 2.52.1, < 2.51.3, < 2.50.2
Exploit Status:
no public exploitCVE-2026-113408
CVSS 0Git does not validate a server-advertised HEAD symref target when cloning an empty repository via protocol v2, allowing a malicious server to supply a path-traversal component that causes writes outside the intended reference directory.
Affected Products:
Git SCM Git – < 2.56.1, < 2.55.1, < 2.54.1, < 2.53.1, < 2.52.1, < 2.51.3, < 2.50.2
Exploit Status:
no public exploitCVE-2026-114467
CVSS 0Git assumes at most two GPG signatures when stripping signatures from a message; a message with more than two signatures causes Git to write past the end of a fixed-size stack array, leading to memory corruption and potential denial of service.
Affected Products:
Git SCM Git – < 2.56.1, < 2.55.1, < 2.54.1, < 2.53.1, < 2.52.1, < 2.51.3, < 2.50.2
Exploit Status:
no public exploitCVE-2026-114469
CVSS 0Git's binary patch application code checks only the compressed patch size and not the declared uncompressed size, allowing a crafted binary patch to bypass the 1 GiB input-size check and trigger an out-of-bounds read that crashes Git.
Affected Products:
Git SCM Git – < 2.56.1, < 2.55.1, < 2.54.1, < 2.53.1, < 2.52.1, < 2.51.3, < 2.50.2
Exploit Status:
no public exploitCVE-2026-105437
CVSS 7.5An integer overflow in Git's merge machinery output-size calculation can cause allocation of an undersized buffer, leading to a heap write-out-of-bounds that corrupts memory when performing large merges, including those involving many conflicts and large conflict-marker-size attributes.
Affected Products:
Git SCM Git – < 2.56.1, < 2.55.1, < 2.54.1, < 2.53.1, < 2.52.1, < 2.51.3, < 2.50.2
Exploit Status:
no public exploitCVE-2026-114466
CVSS 0Three git fast-import stream features (rewrite-submodules-from, rewrite-submodules-to, and export-pack-edges) lacked restrictions when processing untrusted input streams, potentially exposing arbitrary file contents via error messages or allowing creation and appending of files outside the repository.
Affected Products:
Git SCM Git – < 2.56.1, < 2.55.1, < 2.54.1, < 2.53.1, < 2.52.1, < 2.51.3, < 2.50.2
Exploit Status:
no public exploit
MITRE ATT&CK® Techniques
Compromise Software Dependencies and Development Tools
Exploitation for Client Execution
Executable Installer File Permissions Weakness
Command and Scripting Interpreter
File and Directory Discovery
Stored Data Manipulation
Application or System Exploitation
Credentials In Files
Potential Compliance Exposure
Mapping incident impact across multiple compliance frameworks.
PCI DSS 4.0 – All System Components Are Protected from Known Vulnerabilities
Control ID: Requirement 6.3.3
NYDFS 23 NYCRR 500 – Access Privileges and Vulnerability Management
Control ID: Section 500.7
DORA (Digital Operational Resilience Act) – ICT Vulnerability and Patch Management
Control ID: Article 10 — ICT-Related Incident Management
CISA ZTMM 2.0 – Software Supply Chain Integrity and Secure Development Controls
Control ID: Application Pillar — Application Security (Optimize Stage)
NIS2 Directive – Security in Network and Information Systems Supply Chains
Control ID: Article 21(2)(e) — Supply Chain Security
ISO/IEC 27001:2022 – Management of Technical Vulnerabilities
Control ID: Annex A 8.8
Sector Implications
Industry-specific impact of the vulnerabilities, including operational, regulatory, and cloud security risks.
Computer Software/Engineering
Git supply-chain vulnerabilities enabling arbitrary code execution during recursive clones directly threaten software engineering pipelines, CI/CD automation, and developer toolchains relying on unpatched Git versions.
Information Technology/IT
IT organizations running GitHub Enterprise Server or Git-dependent automation face heap corruption and path traversal risks, requiring urgent patching to prevent lateral movement and exfiltration across managed infrastructure.
Computer/Network Security
Security tooling and platforms embedding Git internally expose egress and privilege escalation vectors via CVE-2026-111760 and CVE-2026-114466, undermining Zero Trust segmentation and compliance with NIST.800-53.AC-6.
Financial Services
Financial institutions using Git-based DevSecOps pipelines risk supply-chain compromise through memory corruption and fast-import stream exploits, violating PCI.4.0 egress and visibility controls mapped in validated_info.
Sources
- Git security vulnerabilities announcedhttps://github.blog/open-source/git/git-security-vulnerabilities-announced-7/Verified
- Git Official Website - Download and Release Noteshttps://git-scm.com/downloadsVerified
- NVD - National Vulnerability Databasehttps://nvd.nist.gov/Verified
- CISA Known Exploited Vulnerabilities Cataloghttps://www.cisa.gov/known-exploited-vulnerabilities-catalogVerified
- GitHub Enterprise Server Security Advisorieshttps://docs.github.com/en/enterprise-server/admin/release-notesVerified
- MITRE CVE Programhttps://www.cve.org/Verified
Frequently Asked Questions
Cloud Native Security Fabric Mitigations and ControlsCNSF
Based on the attack progression modeled above, these are the defensive controls that would constrain each stage.
Aviatrix Zero Trust CNSF is relevant here because the attack chain depends on build environments reaching untrusted external Git servers and then pivoting freely into cloud workloads — both reachability patterns that segmented, policy-enforced fabric would likely constrain. By limiting outbound paths from CI/CD agents and enforcing workload-level identity boundaries, the attacker's ability to escalate, move laterally, and exfiltrate across cloud environments would likely be significantly reduced.
Control: Cloud Native Security Fabric (CNSF)
Mitigation: Inline enforcement and real-time traffic inspection at the network perimeter would likely reduce the build environment's reachability to untrusted external Git hosts, making it harder for the malicious clone operation to complete without triggering anomaly detection or policy violation.
Control: Zero Trust Segmentation
Mitigation: Identity-based, least-privilege segmentation policies applied to build agents would likely constrain the scope of credentials accessible from a compromised hook execution context, reducing the attacker's ability to assume higher-privilege IAM roles or reach sensitive credential stores from within the build environment.
Control: East-West Traffic Security
Mitigation: East-west enforcement monitoring workload-to-workload and service-to-service traffic flows would likely reduce the attacker's ability to move freely between build pipeline environments and production cloud workloads, constraining lateral reachability across namespace and account boundaries.
Control: Multicloud Visibility & Control
Mitigation: Centralized multicloud traffic observability would likely surface repeated anomalous outbound patterns from build agents — such as beaconing intervals, unexpected FQDN destinations, or malformed protocol sequences — reducing the attacker's ability to maintain undetected C2 communications across cloud boundaries.
Control: Egress Security & Policy Enforcement
Mitigation: FQDN-based egress filtering and outbound traffic policy enforcement on build and cloud workload egress paths would likely constrain unauthorized data transfers to attacker-controlled infrastructure, reducing the range of reachable exfiltration destinations available from within the pipeline environment.
With segmentation, east-west controls, and egress enforcement applied across earlier stages, the attacker's ability to reach artifact publishing endpoints or inject into downstream pipelines at scale would likely be reduced — meaning the blast radius of supply chain compromise may remain more contained than in an unsegmented environment.
Impact at a Glance
Affected Business Functions
- Software Development Pipelines and CI/CD Automation
- Source Code Repository Management and Integrity
- Third-Party and Open-Source Supply Chain Integration
- Hosting Platform Operations (GitHub Enterprise Server / Cloud)
Estimated downtime: 3 days
Estimated loss: N/A
No confirmed data exfiltration has been reported. However, the vulnerabilities create material risk of: (1) arbitrary code execution during recursive clones enabling attacker-controlled hook and configuration injection into developer workstations and CI/CD runners; (2) path traversal enabling writes of attacker-chosen content outside repository boundaries on hosting services; (3) heap and stack memory corruption during merge operations on hosting platforms processing untrusted pull requests, potentially enabling information disclosure or service crashes; (4) arbitrary file read via fast-import error messages exposing sensitive repository or server-side file contents. Affected environments include any organization running unpatched Git versions in automated pipelines, shared hosting services, or developer endpoints that clone or merge untrusted repositories.
Recommended Actions
Key Takeaways & Next Steps
- • Apply Zero Trust Segmentation with identity-based microsegmentation policies to isolate CI/CD build agents from production cloud workloads, ensuring compromised build environments cannot pivot laterally to production namespaces, Kubernetes clusters, or cloud accounts via stolen credentials or kubeconfig files.
- • Deploy Egress Security & Policy Enforcement with FQDN filtering and outbound traffic controls on all build pipeline environments and cloud workloads to block connections to untrusted or unauthorized Git servers, preventing initial recursive clones of malicious repositories and blocking C2 callback channels established by attacker-controlled hooks.
- • Enable Multicloud Visibility & Control with centralized traffic observability across all build, staging, and production environments to detect anomalous automation patterns, repeated malformed Git protocol requests, and suspicious inter-service API calls indicative of post-exploitation lateral movement.
- • Leverage Inline IPS (Suricata) with updated signatures covering known Git exploit traffic patterns (CVE-2026-111760, CVE-2026-114466 fast-import abuse) on egress paths from build infrastructure to identify and block malicious payload delivery and exfiltration attempts through inspectable traffic channels.
- • Enforce Cloud Native Security Fabric (CNSF) distributed inline policy at the pipeline perimeter to validate repository sources, enforce allowlist-based clone policies, and provide real-time inspection of Git protocol traffic, ensuring only trusted repository origins are reachable from automated build and deployment systems.



