acc-check-serialization
Pass
Audited by Gen Agent Trust Hub on Feb 16, 2026
Risk Level: LOW
Full Analysis
- [SAFE] (SAFE): The skill's primary function is to provide heuristic-based code auditing for PHP applications. A thorough review of the instructions and patterns confirms the skill contains no malicious code execution, prompt injection, or obfuscation attempts.
- [INDIRECT_PROMPT_INJECTION] (INFO): The skill is designed to ingest and process external source code, which technically constitutes an injection surface. However, the risk is classified as negligible (INFO) due to the lack of high-privilege capabilities. 1. Ingestion points: PHP source files (via agent file-reading tools). 2. Boundary markers: Not explicitly defined in instructions. 3. Capability inventory: Limited to producing textual analysis reports for the user; no network, file-write, or execution capabilities are granted to this skill. 4. Sanitization: Not applicable as the output is descriptive analysis for human review.
Audit Metadata