boxlang-code-reviewer
Pass
Audited by Gen Agent Trust Hub on Oct 8, 2026
Risk Level: SAFECOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill instructs the agent to execute local commands and scripts (
box run-script,./run) to perform formatting checks, dependency audits, and security scans. This behavior is standard for a developer-oriented skill and relies on the local project environment. - [INDIRECT_PROMPT_INJECTION]: As the skill's primary purpose is to process and review untrusted BoxLang source code, it possesses an inherent attack surface for indirect prompt injection via code comments or string literals.
- Ingestion points: BoxLang source code files provided for review (e.g.,
.bx,.bxs,.bxmfiles). - Boundary markers: The skill does not define specific delimiters for the code input, relying on the agent's natural ability to distinguish code from instructions.
- Capability inventory: The skill encourages text-based analysis and the execution of local scripts (
box,./run) but does not include network operations or external data exfiltration instructions. - Sanitization: No specific sanitization or escaping mechanisms for the processed code are defined in the instructions.
Audit Metadata