clinic-front-desk-audit
Pass
Audited by Gen Agent Trust Hub on Nov 9, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: The skill consists entirely of Markdown instructions and lacks any executable code (scripts, binaries, or configuration files). No automated command execution or tool invocations are requested.
- [PROMPT_INJECTION]: No malicious injection patterns or attempts to override safety guardrails were identified. The 'Rules for the agent' section establishes positive operational constraints to ensure accuracy and limit the agent's scope to the provided data.
- [DATA_EXFILTRATION]: There are no patterns indicating data exfiltration. The instructions focus on processing numeric values provided by the user and explicitly state that patient names or clinical records are not required.
- [EXTERNAL_DOWNLOADS]: The skill includes links to various vendor websites (e.g., heycai.ai, aevaai.com, smith.ai). These are provided solely as informational resources for the end-user within the report format and are not accessed or executed by the agent.
- [INDIRECT_PROMPT_INJECTION]: While the skill processes external input (clinic figures), it includes boundary instructions ('Only the clinic's numbers') and lacks access to risky capabilities (file writing, shell execution), minimizing potential risks.
Audit Metadata