catalyst-smartbrowz
Pass
Audited by Gen Agent Trust Hub on Aug 15, 2026
Risk Level: SAFECOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill documentation provides instructions for using the Catalyst CLI to initialize projects, create Browser Logic functions, and deploy them to the cloud. Specifically, it mentions commands such as
catalyst init,catalyst functions:create, andcatalyst deploy. These are legitimate developer tools provided by the vendor for managing the service. - [PROMPT_INJECTION]: The skill describes functionality for headless browser automation and PDF/Screenshot generation which involves processing external and potentially untrusted data from URLs, raw HTML, and dynamic template inputs. This creates an attack surface for indirect prompt injection where malicious instructions could be embedded in the content being processed. However, this is the primary purpose of the tool and is accompanied by security notes regarding usage on permitted domains.
- Ingestion points: Functions and SDK methods in
references/smartbrowz-basics.md(e.g.,convertToPdf,take_screenshot,generateFromTemplate) accept user-provided URLs and HTML content. - Boundary markers: The provided code snippets do not include specific delimiters or instructions to ignore embedded commands within the processed data.
- Capability inventory: The skill facilitates network operations via browser navigation and the creation of files (PDFs and images) based on external inputs.
- Sanitization: No explicit sanitization or filtering of input content is documented in the implementation examples.
Audit Metadata