k8s-tool
Pass
Audited by Gen Agent Trust Hub on Apr 4, 2026
Risk Level: SAFECOMMAND_EXECUTIONEXTERNAL_DOWNLOADSPROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill enables the execution of arbitrary kubectl commands and shell commands within containerized environments using the bun k8s-tool CLI tool.
- [EXTERNAL_DOWNLOADS]: The skill relies on the @blogic-cz/agent-tools package and references its official GitHub repository for implementation.
- [PROMPT_INJECTION]: The skill presents an indirect prompt injection surface by processing cluster logs and resource metadata. 1. Ingestion points: External data enters through bun k8s-tool logs and describe commands in SKILL.md. 2. Boundary markers: None provided for external data streams. 3. Capability inventory: Includes host-level kubectl execution and pod-level shell execution. 4. Sanitization: No validation or sanitization of retrieved cluster data is mentioned.
Audit Metadata