scraping-html-to-markdown
Pass
Audited by Gen Agent Trust Hub on Aug 22, 2026
Risk Level: SAFECOMMAND_EXECUTION
Full Analysis
- [COMMAND_EXECUTION]: The skill provides numerous examples of how to use the
crawlbergCLI tool. These are standard usage examples for a web scraping utility and do not include dangerous or malicious command patterns. The flags documented (--proxy,--browser-endpoint, etc.) are typical for scraping tools and are presented in a neutral, instructional context. - [INDIRECT_PROMPT_INJECTION]: As a tool designed to ingest external web content and convert it to Markdown for LLM context, it possesses an attack surface for indirect prompt injection (Category 8). The documentation notes the use of
fit_content(an LLM-optimized variant), but does not explicitly detail sanitization routines for instructions embedded in scraped HTML. This is an inherent risk of scraping tasks rather than a malicious defect of the skill itself. - [NO_CODE]: The skill consists entirely of Markdown documentation and shell command examples; it does not ship with standalone scripts, binaries, or complex automation logic.
Audit Metadata