m365-entra-attack
Pass
Audited by Gen Agent Trust Hub on Oct 26, 2026
Risk Level: SAFECOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTIONDYNAMIC_EXECUTION
Full Analysis
- [COMMAND_EXECUTION]: The skill provides instructions for executing external reconnaissance tools like
msftreconto gather tenant information. - [INDIRECT_PROMPT_INJECTION]: The skill processes untrusted data from external sources, specifically Microsoft Entra ID error bodies and tenant discovery outputs. While the skill includes detailed instructions to parse JSON properly and avoid common parsing traps (e.g., the
access_tokensubstring hazard in MFA challenges), these entry points represent a surface for indirect instructions to enter the agent's context. The skill mitigates this through a mandatory human confirmation gate before any active probing or exploitation. - [DYNAMIC_EXECUTION]: Provides a Python script template utilizing the
playwrightlibrary to automate SAML SSO browser flows for credential validation. The script is designed for automated interaction with target service providers and Microsoft login endpoints.
Audit Metadata