go-docker-builds

Pass

Audited by Gen Agent Trust Hub on Nov 8, 2026

Risk Level: SAFEINDIRECT_PROMPT_INJECTIONEXTERNAL_DOWNLOADS
Full Analysis
  • [INDIRECT_PROMPT_INJECTION]: The skill instructs the agent to analyze local files like go.mod and source code to derive configuration for Dockerfiles and GitHub Actions. This pattern introduces a surface where malicious instructions embedded in project metadata could potentially influence the generated build scripts.
  • Ingestion points: The skill requires gathering inputs from go.mod, main.go, and the project source directory to determine Go versions, binary names, and CGO dependencies.
  • Boundary markers: There are no explicit instructions to use delimiters or ignore embedded instructions when reading user-provided source files.
  • Capability inventory: The skill enables the generation of Dockerfile (with shell commands) and CI/CD workflow files (GitHub Actions YAML).
  • Sanitization: No specific validation or escaping mechanisms are described for metadata extracted from the project files.
  • [EXTERNAL_DOWNLOADS]: The skill references several official and well-known external resources for building containers and managing CI workflows.
  • Fetches official base images including golang:1.26-alpine and alpine:3.21 from Docker Hub.
  • Utilizes official GitHub Actions from the actions/ organization for checking out code.
  • Employs official Docker-maintained actions from the docker/ organization for Buildx setup, registry login, and image pushing.
Audit Metadata
Risk Level
SAFE
Analyzed
Nov 8, 2026, 07:30 PM