azure-kubernetes-automatic-readiness
Pass
Audited by Gen Agent Trust Hub on Oct 12, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- Indirect Prompt Injection Surface: The skill evaluates external Kubernetes manifests and Helm charts provided by the user. Processing untrusted external data is an inherent part of the assessment workflow but constitutes a potential surface for indirect prompt injection.
- Ingestion points: Local YAML manifest files, Helm templates, and live cluster workload data accessed via the
mcp_azure_mcp_akstool. - Boundary markers: The instructions define a strict scope for assessment based on specific resource types and include a "No secrets" guardrail to prevent the extraction of sensitive values.
- Capability inventory: The skill utilizes an MCP tool for data discovery and assessment, and it can perform local file writes to apply fixes only after explicit user approval of a presented diff.
- Sanitization: Validation is performed against a structured, versioned constraint specification (
references/constraint-spec-v1.yaml), which focuses on specific technical fields rather than free-form text, reducing the risk of accidental instruction execution.
Audit Metadata