backend-rust

Pass

Audited by Gen Agent Trust Hub on Oct 13, 2026

Risk Level: SAFECOMMAND_EXECUTIONEXTERNAL_DOWNLOADSINDIRECT_PROMPT_INJECTION
Full Analysis
  • [COMMAND_EXECUTION]: The skill provides a suite of Python automation scripts (build.py, test.py, lint.py, audit.py, and check.py) that execute shell commands using subprocess.run(). These scripts are implemented following security best practices by passing arguments as lists rather than shell strings, which prevents common command injection vulnerabilities.\n- [EXTERNAL_DOWNLOADS]: The skill facilitates the setup of standard developer tools by instructing the user to install pre-commit via pip and automatically installing cargo-audit from the official Rust package registry if it is missing. It also configures pre-commit hooks that fetch from well-known repositories such as gitleaks and the official pre-commit-hooks organization on GitHub.\n- [INDIRECT_PROMPT_INJECTION]: The tdd_gate.py script acts as a workflow enforcement hook that reads the agent's tool prompts from sys.stdin to verify the use of test-driven development (TDD) practices. While this involves processing untrusted input to influence tool execution decisions, the implementation uses deterministic regex pattern matching for validation.\n
  • Ingestion points: scripts/tdd_gate.py reads tool input through sys.stdin.\n
  • Boundary markers: None present.\n
  • Capability inventory: The script controls tool execution permissions by outputting structured JSON instructions to the host environment.\n
  • Sanitization: Employs predefined regex patterns to detect TDD indicators or project initialization commands within the prompt text.
Audit Metadata
Risk Level
SAFE
Analyzed
Oct 13, 2026, 08:04 AM