design-spec-extraction
Warn
Audited by Gen Agent Trust Hub on Jun 27, 2026
Risk Level: MEDIUMCOMMAND_EXECUTIONPROMPT_INJECTIONDATA_EXFILTRATION
Full Analysis
- [COMMAND_EXECUTION]: The skill instructs the agent to execute shell commands such as
mkdir -p .tmp-design-specs/{project-name}. Since the{project-name}variable is not sanitized, this presents a command injection risk if a malicious name containing shell metacharacters (e.g.,;,&,|) is provided. - [PROMPT_INJECTION]: As the skill analyzes screenshots and mockups, it is vulnerable to Indirect Prompt Injection. Text or instructions hidden within the processed images could attempt to override the agent's logic or trick it into writing to unauthorized file paths.
- [DATA_EXFILTRATION]: Extracted design data is consolidated and transferred to a ticketing system via the
todos_oneshottool. Users should ensure the target system is authorized to store sensitive design tokens and specifications. - [FILE_SYSTEM_ACCESS]: The architecture relies on writing intermediate JSON files to paths like
.tmp-design-specs/{project}/. This pattern is vulnerable to path traversal attacks if the{project}variable is manipulated to include directory navigation characters (e.g.,../), potentially allowing files to be written outside the intended temporary directory.
Audit Metadata