dojo-andrew-carnegie
Pass
Audited by Gen Agent Trust Hub on May 2, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: The skill does not contain any malicious patterns. It is comprised entirely of instructional Markdown and business philosophy frameworks.
- [PROMPT_INJECTION]: No attempts to bypass safety filters, extract system prompts, or override core agent instructions were detected. The instructional language used is standard for defining persona-driven behavior and routing logic.
- [DATA_EXFILTRATION]: There is no evidence of network activity or access to sensitive local files (e.g., credentials, environment variables, or SSH keys). All file operations are restricted to the skill's own directory.
- [REMOTE_CODE_EXECUTION]: The skill does not include any scripts, package manager configurations, or commands to download and execute external content. It relies solely on local Markdown content for its knowledge base.
Audit Metadata