dingtalk-message
Pass
Audited by Gen Agent Trust Hub on Oct 25, 2026
Risk Level: SAFECOMMAND_EXECUTIONEXTERNAL_DOWNLOADSINDIRECT_PROMPT_INJECTIONDATA_EXFILTRATION
Full Analysis
- [COMMAND_EXECUTION]: The skill provides a command-line script (
scripts/dingtalk.py) that the agent uses to execute network operations for sending messages.\n- [EXTERNAL_DOWNLOADS]: The skill requires therequestslibrary, which is a standard and well-known Python package for HTTP communication.\n- [INDIRECT_PROMPT_INJECTION]: The skill exposes an attack surface for indirect prompt injection because it processes arbitrary content and file paths to send as messages or attachments. \n - Ingestion points: Message content, titles, and local file paths passed to
scripts/dingtalk.py.\n - Boundary markers: None; inputs are not wrapped in delimiters or safety warnings.\n
- Capability inventory: The skill has the capability to read local files and perform network POST requests to
dingtalk.comdomains.\n - Sanitization: No validation or sanitization is performed on the inputs before they are sent to the API.\n- [DATA_EXFILTRATION]: The skill includes a
filecommand that can read local files and upload them to DingTalk media servers. While this is a documented feature for messaging, it creates a risk of data exfiltration if an agent is tricked into processing sensitive local files.
Audit Metadata