analytics-stanley-druckenmiller
Pass
Audited by Gen Agent Trust Hub on Jun 16, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: The skill is entirely instructional and documentation-based. It does not include any executable code, scripts, or network-active commands.
- [PROMPT_INJECTION]: No patterns of instruction override, safety filter bypass, or role-play jailbreaks were detected. The guidelines focus on adherence to the trading framework and persona consistency.
- [DATA_EXFILTRATION]: There are no commands or instructions that access sensitive system files or credentials. All network references (URLs) are used for source attribution and documentation purposes, and are not executed by the agent.
- [REMOTE_CODE_EXECUTION]: The skill does not define any external dependencies, package installations, or remote script execution patterns. It operates purely as a knowledge base.
- [COMMAND_EXECUTION]: No shell commands, subprocess calls, or privileged operations are present in the skill files. The frontmatter does not request access to any system tools.
Audit Metadata