self-evolution-3d
Pass
Audited by Gen Agent Trust Hub on Aug 18, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: No malicious patterns, obfuscation, or security vulnerabilities were identified during the analysis of the skill instructions and supporting files. The skill functions as a structured coaching guide.
- [EXTERNAL_DOWNLOADS]: The skill includes informational links to the author's website (deciqai.com) and official GitHub repository. It also references reputable external sources, such as the University of Pennsylvania's digital library and Harvard University Press, for historical context and source attribution. These links are static and do not trigger automated code execution.
- [COMMAND_EXECUTION]: There is no evidence of shell command execution, subprocess spawning, or use of dynamic context injection patterns.
- [DATA_EXFILTRATION]: The skill does not access sensitive file paths (e.g., credentials, environment files) or attempt to transmit user data to external servers.
- [PROMPT_INJECTION]: The instructions use standard coaching prompts and wait-states to manage conversation flow. No attempts to bypass safety filters, extract system prompts, or override agent behavior were detected.
Audit Metadata