dot-device-openapi
Pass
Audited by Gen Agent Trust Hub on Oct 30, 2026
Risk Level: SAFEEXTERNAL_DOWNLOADSCOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTIONDYNAMIC_EXECUTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill provides scripts that read local files (images, JSON) and transmit them to the vendor's API. This represents an attack surface for indirect prompt injection where a malicious instruction could trick the agent into reading and exfiltrating sensitive local files.
- Ingestion points: CLI arguments in
scripts/send_image.py,scripts/send_text.py, andscripts/send_canvas.pythat accept local file paths. - Boundary markers: No delimiters or warnings are used when processing local file content.
- Capability inventory: File system read access and network transmission capabilities to
https://dot.mindreset.tech. - Sanitization: No validation is performed on the file paths or the content being sent.
- [EXTERNAL_DOWNLOADS]: The skill provides instructions for installing a local MCP server from the vendor's GitHub repository (
github.com/MindReset/dot_skill.git) usingpipxoruvx. - [COMMAND_EXECUTION]: The skill utilizes several local Python scripts to interact with the device API, requiring the agent to execute these scripts in the local environment.
- [DYNAMIC_EXECUTION]: The
scripts/_runtime.pyfile dynamically modifies the Python module search path (sys.path) at runtime to locate its dependencies within a specific directory structure.
Audit Metadata