azure-functions-agents
Pass
Audited by Gen Agent Trust Hub on Aug 9, 2026
Risk Level: SAFEEXTERNAL_DOWNLOADSREMOTE_CODE_EXECUTIONCOMMAND_EXECUTION
Full Analysis
- Remote Resource Integration: The skill retrieves project templates and manifest files from remote servers to scaffold new applications. This behavior is consistent with modern development workflows that pull official starters from a centralized repository.
- Sandboxed Execution Environment: The skill utilizes dynamic sessions for executing Python code and performing browser automation. This allows agents to handle complex data tasks within a managed, isolated environment, which is an intended capability of the platform.
- Credential Management via Tooling: The instructions provide methods for retrieving and using function keys and tokens through official command-line tools. These keys are necessary for authorizing calls to deployed endpoints and managing resources.
- Input Processing Surface: The agents defined by this skill are designed to ingest data from diverse triggers, such as HTTP endpoints and event connectors. The documentation provides guidance on implementing safety boundaries and handling unexpected data shapes to manage the considerations associated with processing external content.
Audit Metadata