design-audit
Pass
Audited by Gen Agent Trust Hub on Jul 3, 2026
Risk Level: SAFEPROMPT_INJECTIONCOMMAND_EXECUTION
Full Analysis
- [PROMPT_INJECTION]: The skill processes untrusted dashboard source code during its audit phase, creating a surface for indirect prompt injection.
- Ingestion points: Dashboard source files (
src/app/(dashboard)*/**/*.tsx) and UI components (src/ui/*.tsx) are read to extract hardcoded strings for copy auditing. - Boundary markers: Absent. The instructions for the LLM-based copy audit do not specify the use of delimiters or instructions to ignore potential commands embedded in the strings being analyzed.
- Capability inventory: The skill has the ability to create and comment on Linear issues via MCP tools (
linear:create_issue,linear:create_comment) and run shell commands (git rev-parse HEAD). - Sanitization: Limited. While the skill specifies that Linear ticket bodies should be plain markdown without HTML, it does not provide explicit sanitization for the codebase content before it is processed by the LLM or interpolated into ticket reports.
- [COMMAND_EXECUTION]: The skill uses the shell to execute
git rev-parse HEAD. This is used to retrieve the current commit SHA for reporting purposes, which is a standard and safe operation for a development-focused agent skill.
Audit Metadata