criticalSupply Chain

Wazuh GitHub Actions Shell Injection via Crafted VERSION.json (CVE-2026-67308)

First seen Aug 2, 2026 · Updated Aug 2, 2026 · CVSS 10

shell-injectiongithub-actionsci-cdsupply-chainsecrets-exfiltrationself-hosted-runnerswazuhagent-relevant

A critical shell injection vulnerability in Wazuh's GitHub Actions workflows allows attackers to execute arbitrary commands by submitting malicious pull requests containing crafted VERSION.json files. Because affected variables are directly interpolated into shell run steps, attackers can achieve command execution and exfiltrate sensitive secrets such as GITHUB_TOKEN and AWS credentials, particularly dangerous on self-hosted runners with broader network and credential access.

Technical Analysis

The vulnerability stems from unsanitized interpolation of untrusted PR-controlled data (VERSION.json contents) into GitHub Actions run: steps, a classic CI/CD shell injection pattern (similar to CWE-78) rather than a parser or memory-safety bug. An attacker submitting a pull request can inject shell metacharacters (e.g., $(), backticks, ; and |) into environment variables that are later expanded unquoted in bash, resulting in arbitrary command execution within the workflow's runner context. On self-hosted runners this is especially severe because such runners often have access to internal networks, deployment credentials, and long-lived secrets (GITHUB_TOKEN, AWS keys) that can be exfiltrated or used for lateral movement and further supply-chain compromise of downstream artifacts. Wazuh is a widely deployed open-source security monitoring/XDR platform, so a compromised build pipeline could lead to tampered releases or agents. Organizations that use Wazuh's agents or SIEM outputs to feed AI-driven security orchestration, RAG-based alert triage, or autonomous SOC agents should treat any compromised build artifacts or leaked CI credentials as a potential vector for poisoning agent tool inputs or exposing API keys those agents depend on, warranting immediate credential rotation and pipeline validation.

Affected Systems

Wazuh GitHub repository workflows prior to commit 44bf114; GitHub Actions CI/CD pipelines using self-hosted runners; any downstream systems relying on GITHUB_TOKEN or AWS credentials scoped to these workflows

Indicators of Compromise

  • VERSION.json (crafted/malicious pull request payload)
  • Commit hash prior to: 44bf114
  • Suspicious PRs modifying VERSION.json with shell metacharacters ($, `, ;, |, &&)

Remediation Steps

  1. 1

    Update to patched workflow version

    Upgrade Wazuh workflow definitions to commit 44bf114 or later, which remediates the shell injection flaw.

  2. 2

    Sanitize and quote workflow inputs

    Ensure all PR-controlled data used in GitHub Actions steps is passed via environment variables with proper quoting, or processed through safe scripting (e.g., using actions/github-script) rather than direct shell interpolation.

  3. 3

    Rotate exposed credentials

    Immediately rotate GITHUB_TOKEN scopes and any AWS credentials accessible to CI workflows that may have been exposed prior to patching.

  4. 4

    Restrict self-hosted runner permissions

    Limit self-hosted runner network access and credential scope, and require manual approval for workflow runs triggered by external pull requests.

  5. 5

    Audit recent workflow runs

    Review CI/CD logs and pull request history for anomalous VERSION.json submissions or unexpected command execution in workflow runs.

  6. 6

    Enable branch protection and PR review gating

    Require maintainer approval before workflows execute on forked/external pull requests to prevent untrusted code from triggering privileged CI jobs.

CVE / Advisory IDs

CVE-2026-67308

Industries Most Exposed

TechnologySoftware DevelopmentManaged Security ServicesCybersecurityCloud Infrastructure

Sources

Respond to this threat

Pro subscribers get a full AI-generated incident-response playbook for this threat — detection, containment, eradication, and recovery steps — plus an unlimited AI Threat Advisor for questions about your environment.