design-documents

Pass

Audited by Gen Agent Trust Hub on Apr 7, 2026

Risk Level: SAFE
Full Analysis
  • [SAFE]: The skill is entirely instructional and template-based. It provides guidelines for problem analysis, goal definition, and decision-making within the context of software architecture.
  • [DATA_EXFILTRATION]: The skill does not perform any network operations. It instructs the agent to save documentation locally in a design-docs/ directory. No sensitive file access patterns or credential harvesting techniques were identified.
  • [REMOTE_CODE_EXECUTION]: There are no scripts, package dependencies, or shell command executions associated with this skill. It does not download or execute external code.
  • [PROMPT_INJECTION]: The instructions are focused on document structure and quality. No attempts to bypass agent constraints, override safety guidelines, or extract system prompts were found.
  • [COMMAND_EXECUTION]: The skill does not utilize any shell tools or execute subprocesses. All operations are confined to reading and writing Markdown documentation.
Audit Metadata
Risk Level
SAFE
Analyzed
Apr 7, 2026, 04:30 PM
Security Audit — agent-trust-hub — design-documents