smart-document-parser
Pass
Audited by Gen Agent Trust Hub on Aug 27, 2026
Risk Level: SAFECOMMAND_EXECUTION
Full Analysis
- [COMMAND_EXECUTION]: The skill uses
curlto interact with the Smart Document REST API (https://connect.zhihuiya.com/rd-llm/v1/documents/doc_parsing). The commands are statically defined and used for the primary purpose of document parsing. - [DATA_EXPOSURE_AND_EXFILTRATION]: The skill documentation explicitly instructs the agent to read the API key from the environment variable
SMART_DOCUMENT_API_KEYand warns against hardcoding, echoing, or saving the key in logs or responses. This is a secure credential management practice. - [EXTERNAL_DOWNLOADS]: The skill communicates with
connect.zhihuiya.com, which is the official production domain for the service provided by the vendor (Patsnap). This is considered a safe vendor resource. - [REMOTE_CODE_EXECUTION]: No remote code execution patterns were detected. The skill uses a specific
curlcommand to send a file to a REST endpoint and receive a JSON response. - [INDIRECT_PROMPT_INJECTION]: The skill processes external documents (PDFs, images). While it lacks explicit boundary markers for the extracted content, the risk is mitigated by instructions not to invent OCR content and the standard nature of the document parsing task.
Audit Metadata