add-sharepoint

Pass

Audited by Gen Agent Trust Hub on Nov 7, 2026

Risk Level: SAFECOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
  • Command Execution via Official Tooling: The skill instructions involve executing commands through the Power Apps CLI (pa) and Azure CLI (az). These tools are essential for authenticating with Microsoft services, discovering SharePoint sites, and configuring data sources within the application environment.\n- Microsoft Graph API Integration: The skill provides PowerShell functions that utilize Invoke-RestMethod to communicate with graph.microsoft.com. These interactions are used for discovering site structures and managing list schemas, which are legitimate operations for the skill's intended purpose.\n- Secure Authentication Patterns: Authentication is handled by retrieving temporary access tokens through the Azure CLI (az account get-access-token), ensuring that no static credentials or secrets are stored within the skill's logic.\n- Indirect Prompt Injection Surface: The process of generating SharePoint schemas based on user requirements represents a potential ingestion point for external data.\n
  • Ingestion points: User-supplied list definitions and site metadata retrieved via Graph API calls.\n
  • Boundary markers: The workflow utilizes interactive tools like AskUserQuestion and project planning modes to ensure user oversight of the agent's proposed actions.\n
  • Capability inventory: The skill possesses the capability to execute shell commands, perform authenticated network requests, and modify local project documentation.\n
  • Sanitization: Data is processed through structured formats (JSON) during API communication, though it is recommended that agents validate the content of user-provided schemas before execution.
Audit Metadata
Risk Level
SAFE
Analyzed
Nov 7, 2026, 08:17 AM