smartlink-ip-workbench
Warn
Audited by Gen Agent Trust Hub on Aug 27, 2026
Risk Level: MEDIUMCOMMAND_EXECUTIONDYNAMIC_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill uses
screeninSKILL.mdto launch background Node.js and Python processes (python3 -m http.server) for local proxies and servers. - [DYNAMIC_EXECUTION]: The script
scripts/verify_html_js.jsutilizes thenew Function()constructor to validate JavaScript syntax by compiling code blocks extracted from local HTML files. - [INDIRECT_PROMPT_INJECTION]: The skill is designed to read and modify local HTML files, creating an attack surface for instructions embedded in those files.
- Ingestion points: Core workflows in
SKILL.mdand theverify_html_js.jsscript read data from local HTML files located in the user's Downloads directory. - Boundary markers: No delimiters or safety instructions are used when ingesting the external HTML content.
- Capability inventory: The skill possesses file-write capabilities (via
save_latest.sh), shell execution capabilities (viascreen), and dynamic JavaScript evaluation. - Sanitization: There is no evidence of sanitization or safety filtering for the content read from local files before it is processed or evaluated.
Audit Metadata