dockerfile-skill

Pass

Audited by Gen Agent Trust Hub on Jun 15, 2026

Risk Level: SAFECOMMAND_EXECUTIONEXTERNAL_DOWNLOADSPROMPT_INJECTION
Full Analysis
  • [COMMAND_EXECUTION]: The skill utilizes shell commands to perform deep project analysis and executes docker buildx build, docker-compose up, and database CLI tools like psql to validate that the generated environment is functional.\n- [EXTERNAL_DOWNLOADS]: The skill incorporates templates that fetch the rustup toolchain and the bun runtime from their official well-known service domains during the build process.\n- [PROMPT_INJECTION]: The skill has an indirect prompt injection surface as it ingests untrusted project files to determine build requirements.\n
  • Ingestion points: Reads files such as package.json, next.config.js, and ORM schema files (e.g., schema.prisma) to extract configuration metadata.\n
  • Boundary markers: Does not employ explicit delimiters or 'ignore' instructions when processing the content of these external project files.\n
  • Capability inventory: Has extensive access to host-level tools including docker, docker-compose, and various language runtimes.\n
  • Sanitization: Does not demonstrate explicit validation or escaping of data extracted from project files before using it in generation logic.
Audit Metadata
Risk Level
SAFE
Analyzed
Jun 15, 2026, 03:13 AM
Security Audit — agent-trust-hub — dockerfile-skill