rust-deployable-service

Pass

Audited by Gen Agent Trust Hub on Jun 21, 2026

Risk Level: SAFECOMMAND_EXECUTIONEXTERNAL_DOWNLOADSPROMPT_INJECTION
Full Analysis
  • [COMMAND_EXECUTION]: The skill includes a preflight script (scripts/docker-preflight.sh) that automates Docker builds and execution. It allows running a smoke test command defined by the RUST_DEPLOY_SMOKE_COMMAND environment variable within the container, which is a standard verification technique for deployment pipelines.
  • [EXTERNAL_DOWNLOADS]: The documentation references official Docker images from the Rust and Debian projects. These are well-known and trusted sources used for creating reproducible build environments.
  • [PROMPT_INJECTION]: The skill instructions involve analyzing local repository files like Cargo.toml and Dockerfile. This represents a surface for indirect prompt injection if the project being analyzed contains malicious instructions. However, this ingestion is necessary for the skill's primary function.
  • Ingestion points: Local project files, including Cargo.toml, Dockerfile, and configuration files (documented in SKILL.md).
  • Boundary markers: Absent.
  • Capability inventory: Docker command execution and container interactions via the docker-preflight.sh script.
  • Sanitization: Absent within the skill instructions; the agent relies on its internal safety guardrails.
  • [SAFE]: The skill provides proactive security guidance, such as recommending the use of SecretString to prevent accidental credential leakage in logs and explicitly advising against committing production secrets to version control.
Audit Metadata
Risk Level
SAFE
Analyzed
Jun 21, 2026, 03:24 PM
Security Audit — agent-trust-hub — rust-deployable-service