engineering-refactor

Pass

Audited by Gen Agent Trust Hub on Oct 13, 2026

Risk Level: SAFEEXTERNAL_DOWNLOADSCOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
  • [EXTERNAL_DOWNLOADS]: The skill instructions include a command to install a necessary companion skill, engineering-rules, from the vendor's own repository (collectiveai-team/botica). This is a standard procedure for extending the agent's capabilities using vendor-provided resources.
  • [COMMAND_EXECUTION]: The refactoring workflow involves executing standard development tools for linting, formatting, and testing, such as ruff, pytest, uv, pyrefly, and ast-grep. These tools are used to verify code changes and ensure compliance with architectural standards.
  • [INDIRECT_PROMPT_INJECTION]: The skill is designed to analyze and modify a project's codebase, which entails reading and processing external source files and configuration documents. This is the core functionality of the skill and represents a standard surface for processing external content within a development environment.
  • Ingestion points: Local repository files including source code, pyproject.toml, and documentation.
  • Boundary markers: None identified.
  • Capability inventory: Execution of development tools and skill installation commands.
  • Sanitization: Not applicable for code refactoring tasks.
Audit Metadata
Risk Level
SAFE
Analyzed
Oct 13, 2026, 07:16 AM