skills/marcioaltoe/skills/rust-cli/Gen Agent Trust Hub

rust-cli

Fail

Audited by Gen Agent Trust Hub on Jul 3, 2026

Risk Level: HIGHREMOTE_CODE_EXECUTIONCOMMAND_EXECUTIONEXTERNAL_DOWNLOADSPROMPT_INJECTION
Full Analysis
  • [REMOTE_CODE_EXECUTION]: The Makefile template in references/templates.md includes a pre-commit-hooks target that executes curl -LsSf https://github.com/j178/prek/releases/latest/download/prek-installer.sh | sh. This pattern is dangerous as it executes unverified code from an external source directly in the user's shell environment.
  • [COMMAND_EXECUTION]: The skill recommends several shell commands for environment setup, including cargo install cargo-release in SKILL.md and both cargo install cargo-release and cargo install prek in references/templates.md.
  • [EXTERNAL_DOWNLOADS]: The skill suggests downloading and installing tools from external repositories and registries (GitHub, crates.io) without providing version pins or integrity checks for the installation process.
  • [PROMPT_INJECTION]: The skill creates a surface for indirect prompt injection by instructing the agent to adapt templates based on user projects without defining boundary markers or sanitization steps for the ingested data.
  • Ingestion points: User-provided project structure and requirements during template adaptation.
  • Boundary markers: None provided in the instructions.
  • Capability inventory: Shell command execution capabilities via the generated Makefile and cargo commands.
  • Sanitization: None specified for the interpolation of project data into templates.
Recommendations
  • AI detected serious security threats
Audit Metadata
Risk Level
HIGH
Analyzed
Jul 3, 2026, 01:21 PM
Security Audit — agent-trust-hub — rust-cli