spring-ai-security
Pass
Audited by Gen Agent Trust Hub on Nov 9, 2026
Risk Level: SAFE
Full Analysis
- [PROMPT_INJECTION]: The skill includes common prompt injection patterns, such as 'Ignore earlier instructions', within its documentation (e.g.,
examples/good-boundary.md). These are explicitly identified as synthetic malicious payloads for the purpose of testing security boundaries and do not constitute an attempt to hijack the agent. - [DATA_EXFILTRATION]: No network exfiltration or credential harvesting patterns were detected. The skill's instructions specifically warn against data exposure and recommend masking sensitive identifiers in logs.
- [INDIRECT_PROMPT_INJECTION]: The skill provides comprehensive guidance on mitigating indirect prompt injection. It identifies ingestion points for untrusted data (user input, documents, retrieved chunks) and recommends the use of server-side authorization, call budgets, and strict boundary enforcement in application code.
- [COMMAND_EXECUTION]: No unauthorized command execution patterns were found. The provided Java template (
templates/DocumentReadTool.java) demonstrates best practices for tool isolation, including type-safety and permission checks before execution.
Audit Metadata