java-file-audit
Pass
Audited by Gen Agent Trust Hub on Sep 23, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill is designed to ingest and analyze untrusted source code and data flow evidence provided by external audit pipelines. This creates a surface for indirect prompt injection if the audited content contains malicious instructions designed to subvert the agent's logic. However, this is inherent to the auditing function and no active exploitation was found.
- Ingestion points: Source code files and evidence data provided by the java-audit-pipeline.
- Boundary markers: The skill does not explicitly define delimiters to separate audited source code from the agent's instructions, though it operates in a documentation/reference capacity.
- Capability inventory: The skill itself contains instructional guidelines and does not trigger subprocesses or network operations.
- Sanitization: No explicit sanitization of the audited code blocks is mentioned.
- [SAFE]: The skill provides high-quality defensive guidance, including path normalization best practices, UUID renaming for uploaded files, and canonical path verification for archive extraction (Zip Slip mitigation). The potentially dangerous code snippets are clearly labeled as educational examples for security researchers.
Audit Metadata