xmind
Pass
Audited by Gen Agent Trust Hub on May 18, 2026
Risk Level: SAFEDATA_EXFILTRATIONEXTERNAL_DOWNLOADSCOMMAND_EXECUTION
Full Analysis
- [DATA_EXFILTRATION]: The skill features an "attachment" field that allows the agent to read any file on the system via an absolute path and embed its contents into the generated .xmind file. If an agent is maliciously directed, this could lead to the exposure of sensitive files (e.g., configuration files or private keys) by including them as mind map attachments.
- [EXTERNAL_DOWNLOADS]: To support mind map generation from large PDF files, the skill instructions include a Python script that dynamically installs the "pymupdf" package from PyPI. While this is a standard library, dynamic installations introduce a dependency on external package registries.
- [COMMAND_EXECUTION]: The skill workflow involves piping JSON data into a Node.js script and running CLI tools like "pdftotext". These operations involve shell execution where file paths are passed as arguments, requiring careful handling of inputs to avoid unintended command behavior.
Audit Metadata