nocobase-ai-employee-manager
Pass
Audited by Gen Agent Trust Hub on Aug 2, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: The skill manages AI employee records using the NocoBase CLI (nb). Its operations are consistent with its stated purpose and focus on local API interactions.
- [COMMAND_EXECUTION]: The skill utilizes the Bash tool to execute nb api commands. It follows security best practices by using protected body files for structured JSON payloads to prevent command injection.
- [PROMPT_INJECTION]: Instructions include strict safety gates, such as mandatory confirmations for high-risk actions like employee deletion or modifying answer sources. It explicitly prohibits bypassing validation steps or modifying protected internal fields. The skill also processes potentially untrusted data such as employee bios and knowledge base content, mitigating risks by requiring specific placeholders ({knowledgeBaseData}) and performing pre-flight validation of capabilities and keys. (Ingestion: API/User input; Boundaries: Prompt placeholders; Capability: Bash, Read; Sanitization: Field rejection and dependency validation).
- [DATA_EXFILTRATION]: No evidence of data exfiltration was found. The skill operates within the authenticated NocoBase CLI environment and does not attempt to send data to external or untrusted domains.
Audit Metadata