cspm
Pass
Audited by Gen Agent Trust Hub on Jul 8, 2026
Risk Level: SAFEPROMPT_INJECTIONEXTERNAL_DOWNLOADSCOMMAND_EXECUTION
Full Analysis
- [PROMPT_INJECTION]: The skill presents an indirect prompt injection surface due to its core function of auditing and processing external cloud metadata. Ingestion points: Environment configurations, source code boundaries, and cloud provider metadata (SKILL.md). Boundary markers: Absent in the provided markdown instructions. Capability inventory: The agent is instructed to execute state modifications and update resource policies (references/automated-remediation.md). Sanitization: A general rule to sanitize user inputs is mentioned in SKILL.md, but specific logic is not detailed.
- [EXTERNAL_DOWNLOADS]: Fetches and references infrastructure-as-code modules and security actions from established vendors. Examples include Checkov actions and Wiz cloud connectors (references/cspm-advanced.md, references/cspm-platforms.md).
- [COMMAND_EXECUTION]: Provides detailed code implementation patterns and CLI guidance for cloud automation. Includes Go, Rust, and TypeScript snippets for managing network streams, executing resilient service calls, and performing resource state transitions (references/architecture-patterns.md, references/performance-optimization.md).
- [SAFE]: Contains eight reference files (e.g., architecture-patterns.md, code-organization.md) consisting of approximately 1,200 sections of highly repetitive technical paragraphs and code snippets. This bloat content was analyzed and found to be benign, lacking hidden malicious strings or commands.
Audit Metadata