init-workspace

Pass

Audited by Gen Agent Trust Hub on Aug 3, 2026

Risk Level: SAFEEXTERNAL_DOWNLOADSCOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
  • [EXTERNAL_DOWNLOADS]: Downloads documentation templates from the author's official GitHub repository (gtrabanco/agentic-workflow/template) using the npx degit utility.
  • [COMMAND_EXECUTION]: Utilizes the GitHub CLI (gh) to automatically seed triage labels ('urgent' and 'fix-next') within the repository, subject to user approval.
  • [EXTERNAL_DOWNLOADS]: Suggests the installation of several well-known third-party development tools and plugins (e.g., Vitest, Biome, ESLint plugins) based on the project's detected technology stack.
  • [PROMPT_INJECTION]: Analyzed for indirect prompt injection potential through the processing of project manifest files and documentation.
  • Ingestion points: Processes local files such as package.json, pyproject.toml, go.mod, Cargo.toml, Gemfile, and CLAUDE.md to detect project settings.
  • Boundary markers: Does not implement specific delimiters when reading content from these files for configuration discovery.
  • Capability inventory: Includes file system modification (writes and folder pruning), repository state management, and package installation via npx.
  • Sanitization: Mitigates risk by presenting all discovered information as proposals that require explicit user confirmation before any action is executed.
Audit Metadata
Risk Level
SAFE
Analyzed
Aug 3, 2026, 10:34 PM
Security Audit — agent-trust-hub — init-workspace