bx-sites-getting-started

Pass

Audited by Gen Agent Trust Hub on Aug 28, 2026

Risk Level: SAFEREMOTE_CODE_EXECUTIONEXTERNAL_DOWNLOADSCOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
  • [REMOTE_CODE_EXECUTION]: The documentation instructs users to fetch and execute shell scripts from install.boxlang.io and install-bvm.boxlang.io by piping the output of curl to bash. These domains serve the installation logic for the BoxLang runtime environment and its version manager, which are requirements for the BxSites tool.
  • [EXTERNAL_DOWNLOADS]: The skill uses package management tools like CommandBox (box) and the install-bx-module binary to download the bx-sites module and its dependencies from distribution registries.
  • [INDIRECT_PROMPT_INJECTION]: The bxSites migrate command introduces a surface for indirect prompt injection by ingesting content from external formats like GitBook, mkdocs, Notion, and ZIP archives.
  • Ingestion points: The migrate command in SKILL.md reads directory structures and Markdown files from user-specified external sources.
  • Boundary markers: The instructions do not specify the use of delimiters or specific markers to ignore embedded instructions within the processed content.
  • Capability inventory: The skill can execute CLI commands, create or modify files in the project directory, and serve content locally on port 8080.
  • Sanitization: No explicit sanitization or validation of the ingested external content is described in the documentation.
  • [COMMAND_EXECUTION]: The skill invokes the bxSites CLI for several operations, including scaffolding new projects (new), creating pages (page:new), and managing the build lifecycle (build, serve, clean). These actions involve reading from and writing to the local file system and opening a local network port for previewing the site.
Audit Metadata
Risk Level
SAFE
Analyzed
Aug 28, 2026, 03:14 AM
Security Audit — agent-trust-hub — bx-sites-getting-started