outlook-emails-lawvable
Pass
Audited by Gen Agent Trust Hub on Apr 2, 2026
Risk Level: SAFEPROMPT_INJECTIONEXTERNAL_DOWNLOADS
Full Analysis
- [PROMPT_INJECTION]: The skill has an indirect prompt injection surface as it processes external email data.
- Ingestion points: The
get_emailsfunction inscripts/outlook_oauth.pyretrieves untrusted data including email subjects, sender names, and body previews. - Boundary markers: There are no delimiters or markers used to isolate the email content from the agent's instructions.
- Capability inventory: The skill can write files to the local filesystem (saving attachments to
./attachments) and perform authorized network requests to the Microsoft Graph API. - Sanitization: While the script implements path traversal protection for attachment filenames, it does not sanitize or escape the content of the email body or subject.
- [EXTERNAL_DOWNLOADS]: The skill uses the
uvpackage manager to download and install standard Python dependencies includingmsal,requests, andpython-dotenvfrom the official Python Package Index (PyPI).
Audit Metadata