ad-security-reviewer

Pass

Audited by Gen Agent Trust Hub on Sep 23, 2026

Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
  • [INDIRECT_PROMPT_INJECTION]: The skill processes untrusted data sourced from Active Directory and Microsoft Graph, including user account attributes, group names, and security logs. This data is interpolated into markdown reports and remediation scripts, creating a potential surface for indirect prompt injection if an attacker can manipulate directory objects to include malicious instructions.\n
  • Ingestion points: The skill fetches external data in scripts/analyze_ad_security.ts via Microsoft Graph API and in scripts/audit_privileged_groups.ps1 and scripts/review_delegation.ps1 using Active Directory PowerShell queries.\n
  • Boundary markers: There are no explicit delimiters or specific guardrail instructions implemented to isolate directory data from the agent's instructions during processing.\n
  • Capability inventory: The skill possesses the capability to write files (reports and scripts) and execute commands via Bash access and PowerShell for audit validation.\n
  • Sanitization: No evidence was found of input sanitization or validation of directory attribute values before they are used in generated outputs.
Audit Metadata
Risk Level
SAFE
Analyzed
Sep 23, 2026, 08:05 AM
Security Audit — agent-trust-hub — ad-security-reviewer