rhdh-base-images

Pass

Audited by Gen Agent Trust Hub on Sep 8, 2026

Risk Level: SAFEREMOTE_CODE_EXECUTIONEXTERNAL_DOWNLOADSCOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
  • [REMOTE_CODE_EXECUTION]: The skill downloads and executes maintenance scripts from gitlab.cee.redhat.com to facilitate base image updates.\n
  • Evidence: Scripts such as updateBaseImages.sh and getLatestImageTags.sh are fetched and run locally.\n- [EXTERNAL_DOWNLOADS]: The skill fetches Node.js headers from nodejs.org and installs the rpm-lockfile-prototype tool from a public GitHub repository.\n
  • Evidence: Tooling is installed via pip from the konflux-ci organization.\n- [COMMAND_EXECUTION]: The skill leverages various system and DevOps utilities including git, gh, podman, docker, skopeo, and python3 for repository maintenance and container management.\n- [INDIRECT_PROMPT_INJECTION]: The skill processes project-specific files such as Containerfiles and Dockerfiles, creating a potential surface for instructions embedded in data to influence execution.\n
  • Ingestion points: Containerfiles, Dockerfiles, and YAML configuration files in the target repositories.\n
  • Boundary markers: No explicit boundary delimiters are used during parsing.\n
  • Capability inventory: File system write access, network operations via curl and git, and shell command execution.\n
  • Sanitization: Data is processed using standard text utilities (grep, sed) without specific sanitization for malicious instructions.
Audit Metadata
Risk Level
SAFE
Analyzed
Sep 8, 2026, 04:32 PM
Security Audit — agent-trust-hub — rhdh-base-images