space-brand-ui-design
Pass
Audited by Gen Agent Trust Hub on Jul 13, 2026
Risk Level: SAFEEXTERNAL_DOWNLOADSREMOTE_CODE_EXECUTIONCOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill instructs the agent to execute multiple shell commands to manage temporary directories and retrieve data.
- Evidence: SKILL.md line 125: 'cd /tmp && mkdir -p design-md-tmp && cd design-md-tmp && npx getdesign@latest add 2>&1'
- [REMOTE_CODE_EXECUTION]: The skill uses 'npx' with the '@latest' tag, which downloads and executes the most recent version of an external package from the NPM registry at runtime.
- Evidence: SKILL.md line 125: 'npx getdesign@latest add '
- [EXTERNAL_DOWNLOADS]: The skill fetches design specifications ('DESIGN.md') from an external source via the 'getdesign' tool.
- Evidence: README.md and SKILL.md describe retrieval of specs from getdesign.md.
- [PROMPT_INJECTION]: The skill processes untrusted data from an external file which could be manipulated to influence agent behavior (Indirect Prompt Injection).
- Ingestion points: The agent reads the contents of '/tmp/design-md-tmp/DESIGN.md' (SKILL.md line 129).
- Boundary markers: Absent; the instruction does not specify delimiters or warnings to ignore instructions embedded within the fetched specification.
- Capability inventory: The agent has shell access and is performing code generation tasks based on the input.
- Sanitization: Absent; the skill lacks a mechanism to validate or sanitize the markdown content fetched from the external source before it enters the agent's context.
Audit Metadata