dropbox-engineer
Pass
Audited by Gen Agent Trust Hub on Mar 31, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: No security issues were identified in the analyzed skill files. The content is professionally structured as an expert persona for an AI agent, focusing on technical documentation and educational code implementations related to Dropbox's infrastructure.
- [COMMAND_EXECUTION]: Code examples provided in SKILL.md and reference files are conceptual (Python and Rust) and intended for illustrative purposes. They do not contain instructions to execute arbitrary shell commands or perform unsafe operations on the host system.
- [DATA_EXFILTRATION]: No network exfiltration patterns or hardcoded credentials were found. The network operations described in the examples (e.g.,
_upload_blocks,server_request) are part of conceptual sync engine logic and do not target external or malicious domains. - [PROMPT_INJECTION]: The system prompt correctly defines a professional identity and decision framework without attempting to override the underlying agent's safety filters or core instructions.
Audit Metadata