setup-datamodel
Pass
Audited by Gen Agent Trust Hub on Sep 15, 2026
Risk Level: SAFECOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
- Command Execution: The skill utilizes shell commands and local Node.js scripts to perform environment resolution and schema reconciliation.
- Evidence includes the execution of
resolve-environment.jsandoffline-profile-delta.jslocated within the plugin root. - The skill invokes
npx power-appsto add data sources, which is a standard procedure for this ecosystem. - Indirect Prompt Injection: The skill accepts user-provided input, including diagrams and text descriptions, to inform the data model design.
- This data is ingested during Phase 2 and passed to the
data-model-architectagent. - While this creates an interface for external data, it is used specifically for generating structured schema designs and is subject to agent safety protocols.
Audit Metadata