dhdna-profiler
Pass
Audited by Gen Agent Trust Hub on Jun 26, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: No security concerns were identified. The skill performs natural language analysis on user-provided text to generate cognitive profiles.
- [DATA_EXPOSURE_AND_EXFILTRATION]: All external links point to legitimate research repositories (Zenodo/DOI) or the vendor's own domains (ahkstrategies.net, themindbook.app). There is no evidence of unauthorized data exfiltration or credential harvesting.
- [INDIRECT_PROMPT_INJECTION]: While the skill ingests untrusted text for profiling, its capability is limited to text analysis and formatting within the agent's context. It lacks high-risk capabilities like arbitrary command execution or network exfiltration, making the indirect injection surface non-critical.
Audit Metadata