ui-ux-design
Pass
Audited by Gen Agent Trust Hub on Sep 15, 2026
Risk Level: SAFECOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill provides setup instructions for installing Python on various operating systems using official package managers, including
brew,winget, andapt. The Linux installation instructions include the commandsudo apt install python3to resolve prerequisites for running the search tool. - [INDIRECT_PROMPT_INJECTION]: The skill implements a workflow where user-supplied keywords are used to query a local database of UI/UX guidelines, which the agent is then instructed to follow during design implementation.
- Ingestion points: User keywords are accepted as command-line arguments in
scripts/search.py(e.g.,python3 search.py "<keyword>"). - Boundary markers: The skill does not define explicit delimiters or instructions to ignore embedded commands for the data returned from the CSV files.
- Capability inventory: The skill performs local file reads of CSV data and executes a local Python search script.
- Sanitization: The
BM25.tokenizefunction inscripts/core.pysanitizes input by applying a regular expression (re.sub(r'[^\w\s]', ' ', ... )) to the query string, effectively removing punctuation and special characters that could be used for injection before the search is executed.
Audit Metadata