adf-master

MEDIUM W011: Third-party content exposure detected (indirect prompt injection risk).

• Third-party content exposure detected (high risk: 0.90). This skill explicitly instructs the agent to fetch and use open/public third‑party content (e.g., raw.githubusercontent.com script downloads, Context7 library fetches, WebFetch of Microsoft Learn pages, npm, Medium and community blogs/StackOverflow) which are untrusted or user‑generated sources that the agent is expected to read and interpret at runtime.

MEDIUM W012: Unverifiable external dependency detected (runtime URL that controls agent).

• Potentially malicious external URL detected (high risk: 1.00). The skill includes explicit runtime download-and-execute commands (curl / Invoke-WebRequest) for the PowerShell script at https://raw.githubusercontent.com/Azure/Azure-DataFactory/main/SamplesV2/ContinuousIntegrationAndDelivery/PrePostDeploymentScript.Ver2.ps1, which is fetched at runtime and intended to be executed as part of the deployment flow, so it constitutes a required external script that executes remote code.