auth-tool-cloudbase
Fail
Audited by Gen Agent Trust Hub on Feb 16, 2026
Risk Level: HIGHPROMPT_INJECTION
Full Analysis
- [Prompt Injection] (HIGH): Vulnerable to Indirect Prompt Injection (Category 8) due to sensitive write capabilities on authentication infrastructure. 1. Ingestion points: Variable inputs for EnvId, AppSecret, ClientSecret, and SMTP details in SKILL.md. 2. Boundary markers: Absent; no instructions to isolate or ignore instructions within input data. 3. Capability inventory: High-impact API calls (ModifyLoginStrategy, ModifyProvider) via callCloudApi to tcb and lowcode services across all configuration scenarios. 4. Sanitization: Absent; data is used directly in configuration payloads without verification.
- [No Code] (INFO): Contains markdown instructions and templates only; no executable scripts or standalone code provided.
Recommendations
- AI detected serious security threats
Audit Metadata