Urgent.News

What's breaking now, across thousands of outlets.

Tech

Deleting a secret from your Docker image doesn't delete it from your build history

An autonomous hacking agent found a live GitHub admin token in Baseten's infrastructure by pulling a public Docker image and reading its build history. The token was from March 2023, still had repo scope and admin on internal repos in July 2026. The writeup is a good read on its own , but the mechanism it exposes is the part worth internalizing. The token wasn't in the filesystem layers. It was…

An autonomous hacking agent uncovered a live GitHub admin token in a public Docker image by examining its build history. The token, from March 2023, retained admin permissions on internal repositories as late as July 2026. The token was not stored in the image's filesystem layers but instead in the image's configuration, where a RUN command had directly inserted the GITHUB_TOKEN into history[].created_by.

Docker archives a record of every build step, including this information, alongside the image blobs. Despite scrubbing the secret and rebuilding the image, the old build history still contains the plaintext token. This poses a significant threat as cleanup practices typically focus on individual files, but the real danger lies in metadata.

Credentials passed through RUN or ARG during the build process can become immortal in the image's history unless rebuilt from scratch and never injected in the first place. To address this, it's crucial to consider image build history as part of the audit surface, similar to the layers. Secret scanners often overlook this metadata.

The best approach for a leaked build token is to rotate it and rebuild the image, ensuring the secret is never passed as a build argument again. However, the issue doesn't end there. If images are handed to build agents or pushed to a registry, they carry the built-up history, including potentially dormant credentials. Container security isn't just about containment but also about managing metadata.

Strix and Baseten responded quickly and effectively, emphasizing the importance of this issue. It's worth dedicating around 20 minutes to understand the implications, as the lesson is relatively simple, but the consequences can be severe once your build pipeline is compromised.

Written by urgent.news from Dev.to's reporting — not their text. Machine-written — may contain errors; check the original before relying on it.

Read the original at dev.to →

More in Tech

Yokhash Meets VPC: Exploring Amazon Virtual Private Cloud ☁️

📝 Introduction What is the AWS service? Amazon VPC (Virtual Private Cloud) is a networking service that lets you create your own logically isolated section of the AWS cloud.

  • Amazon Virtual Private Cloud (VPC) enables isolated AWS cloud sections.
  • VPC launched in 2009 to improve secure, enterprise-grade architectures.
  • VPC provides control over IP range, subnets, route tables, and gateways.

OpenJAI-v1.0-14B, a Thai model from the same team as JaiTTS, made available for free.

OpenJAI-v1.0-14B โมเดลไทยจากทีมเดียวกับ JaiTTS ที่เปิดฟรี โดย Nokka (นก-กา) | 15 กันยายน 2026 บทความนี้เขียนโดย AI (โมเดล deepseek-v4.1-flash ของผู้ให้บริการ ollama-cloud) ผ่าน Hermes Agent จาก Nous…

  • OpenJAI-v1.0-14B is an open-source Thai language model developed by Nokka.
  • Model outperforms Qwen3-14B in 7 out of 10 benchmarks.
  • Model remains stable in knowledge base after fine-tuning.

#M for Midhun, M for Macie: Exploring Amazon Macie 🔐☁️

Introduction Cloud storage makes it easy for organizations, colleges, and students to store large amounts of data. However, storing data in the cloud also creates an important question: How do we know…

  • Amazon Macie discovers sensitive data in S3 using ML and pattern matching
  • Identifies PII, financial info, credentials with AWS or custom identifiers
  • Provides automated discovery and discovery jobs for security monitoring

More from Wednesday 16 September →