mobbin-search
Pass
Audited by Gen Agent Trust Hub on Sep 15, 2026
Risk Level: SAFECOMMAND_EXECUTIONDYNAMIC_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill instructs the agent to execute the system 'open' command on a file path generated at runtime. The filename is derived from the user's query ('slug'), which could lead to command injection if the agent fails to sanitize special characters before passing the path to the shell.
- [DYNAMIC_EXECUTION]: The skill dynamically generates a self-contained HTML evidence board at runtime, incorporating external image assets and metadata, and writes it to the local directory './.mobbin/'.
- [INDIRECT_PROMPT_INJECTION]: The skill processes data from an external API (Mobbin), which is then used for visual analysis and report generation, creating a vulnerability surface for untrusted data.
- Ingestion points: External metadata and image content enter via the 'search_screens' tool output.
- Boundary markers: The instructions do not specify any delimiters or safety markers for the tool's output data.
- Capability inventory: The skill involves writing files to the local filesystem and executing system commands ('open').
- Sanitization: The skill lacks explicit instructions for sanitizing or escaping ingested strings like 'app_name' before their inclusion in the generated HTML artifacts.
Audit Metadata