skills/uipath/skills/uipath-platform/Gen Agent Trust Hub

uipath-platform

Pass

Audited by Gen Agent Trust Hub on Sep 14, 2026

Risk Level: SAFEDATA_EXFILTRATIONEXTERNAL_DOWNLOADSDYNAMIC_CONTEXT_INJECTIONINDIRECT_PROMPT_INJECTION
Full Analysis
  • [DATA_EXFILTRATION]: The skill identifies the filesystem path for the UiPath authentication token (~/.uipath/.auth) to enable fallback REST API calls. This usage is confined to the skill's primary purpose of platform interaction and does not involve unauthorized exfiltration.
  • [EXTERNAL_DOWNLOADS]: The skill references the installation of official CLI extensions, specifically @uipath/data_fabric_tool, using the tool's built-in manager. These are verified vendor-provided resources.
  • [DYNAMIC_CONTEXT_INJECTION]: Documentation describes the !command syntax used by certain agent platforms to facilitate user-driven interactive login. This is provided for educational purposes and does not represent an automated load-time execution risk.
  • [INDIRECT_PROMPT_INJECTION]: The skill processes data from external entities and services. To mitigate the resulting attack surface, it implements a comprehensive safety framework including mandatory user-approval gates (Rule 0) and schema-preview cycles (Rule 14) for all destructive or schema-altering operations.
Audit Metadata
Risk Level
SAFE
Analyzed
Sep 14, 2026, 06:04 PM
Security Audit — agent-trust-hub — uipath-platform