{
  "id": 21097399,
  "title": "GitHub Advanced Security alerts: prioritise with runtime context",
  "url": "https://urgent.news/2026/11/21/github-advanced-security-alerts-prioritise-with-runtime-context",
  "topic": "tech",
  "section": "Tech",
  "published": "2026-11-21T19:30:18.000Z",
  "source": {
    "name": "Dev.to",
    "slug": "dev-to",
    "url": "https://dev.to/spanobserver/github-advanced-security-alerts-prioritise-with-runtime-context-322h"
  },
  "original_language": "en",
  "account": "On Monday morning, the security tab of a GitHub organization displays thousands of Dependabot and code scanning findings. However, not all of them can be fixed within a week, leading to the operational question of which ones should be addressed first. GitHub Advanced Security alerts indicate which code is vulnerable but do not provide information on whether the affected image is running or reachable from the internet. Dynatrace can provide this missing context for Kubernetes workloads. This article follows the order in which questions should be asked: what is needed, how an image is linked to its repository, how to verify data flow, and how to filter within GitHub. For an overview of the feature, refer to Dynatrace Application Security first.\n\nThe security tab initially shows an overwhelming number of alerts, making it difficult to prioritize which ones to address first. Without context, every alert appears equally urgent. The solution is to provide deployment information, as alerts originating from services exposed in production are more critical than those from libraries never deployed. Dynatrace can identify which container images are actually running and match them to their respective GitHub repositories. This information is then recorded and includes details on internet exposure and sensitive data reachability if Runtime Vulnerability Analytics (RVA) is active.\n\nFor Dynatrace SaaS, this requires an ActiveGate version 1.310 or later that can run Extensions 2.0, reach the GitHub API, and connect to Grail with appropriate permissions. A platform token with storage:smartscape:read, storage:buckets:read, and storage:security.events:read permissions is necessary for runtime context. On GitHub, using a GitHub App is recommended over a personal access token due to its fine-grained permissions, organization-level audit logs, and higher API limits. The app must be installed on each organization to be monitored, and it enables Dynatrace to write deployment records and read attestations. Image-to-repository linking can be achieved through image name (the fallback, which only works for images hosted on ghcr.io) or an attestation in the build workflow. Once the data is available, GitHub alerts can be filtered to prioritize runtime context, focusing on vulnerabilities in internet-exposed images or those reaching sensitive data.\n\nImplementing this process in GitHub reduces noise and improves the time taken to fix issues, as teams are directed to the most critical problems. However, there are limitations to consider, such as the need for scan completion before ingesting findings, potential issues with component version matching, and the fact that secret scanning alerts default to HIGH risk. For optimal results, start with a pilot organization and install the GitHub App only on relevant repositories, begin with attestation before deployment, and plan for token renewal if the GitHub App is not used.",
  "summary": "On Monday morning the security tab of your GitHub organisation shows thousands of Dependabot and code scanning findings. Nobody can fix them all this week, so the operational question is: which ones first? GitHub Advanced Security alerts tell you what is vulnerable in the code. They do not tell you whether the affected image is running anywhere, or reachable from the internet. Dynatrace can fill…",
  "key_points": [
    "GitHub alerts overwhelm security tab with thousands of findings",
    "Dynatrace provides runtime context to prioritize critical vulnerabilities",
    "Implementing process reduces noise and improves issue resolution time"
  ],
  "editors_take": null,
  "illustration": null,
  "coverage": {
    "outlets": 1,
    "also_reported_by": []
  },
  "ai_generated": true,
  "disclaimer": "Summaries, key points and the editor’s take are written by software from other outlets’ reporting and may contain errors — always check the linked original."
}