attending-physician
Pass
Audited by Gen Agent Trust Hub on Mar 31, 2026
Risk Level: SAFE
Full Analysis
- [PROMPT_INJECTION]: The skill implements a board-certified physician persona with structured decision gates and mental models. The instructions are focused on educational and clinical accuracy and do not attempt to override the model's core safety guidelines or ignore system instructions.\n- [EXTERNAL_DOWNLOADS]: The skill provides references to reputable medical guidelines (ACP, NICE, WHO) and its own source code repository on GitHub for documentation and installation. These are informational and do not involve the execution of untrusted remote code.\n- [COMMAND_EXECUTION]: The documentation includes shell commands (e.g.,
echo ... >> ~/.claude/CLAUDE.md) intended for the user to configure the skill's persistence. These are standard setup procedures for AI agent skills and are not triggered automatically or maliciously by the agent itself.
Audit Metadata