public-space-design
Pass
Audited by Gen Agent Trust Hub on Aug 3, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: The skill consists entirely of instructional text and reference documentation for urban design. No security issues were detected across all analyzed categories.
- [PROMPT_INJECTION]: The instructions strictly define a professional persona and workflow for public space design. There are no attempts to override safety filters, extract system prompts, or bypass agent constraints.
- [DATA_EXFILTRATION]: No network operations, hardcoded credentials, or access to sensitive local file paths were found. The skill only processes user-provided site information to generate a design response.
- [REMOTE_CODE_EXECUTION]: The skill does not download or execute external scripts, nor does it use any package managers or dynamic code execution functions.
- [INDIRECT_PROMPT_INJECTION]: While the skill ingests untrusted user data (site constraints, user needs), it possesses no dangerous capabilities such as file-writing, network access, or shell execution. This lack of capability eliminates the risk of indirect prompt injection attacks.
- Ingestion points: User input via chat during the design workflow (site analysis and needs assessment).
- Boundary markers: Absent, as the skill functions as a template-based text generator.
- Capability inventory: None. No tools or scripts are present.
- Sanitization: Not applicable as there are no executable outputs.
Audit Metadata