linkerd-patterns
Pass
Audited by Gen Agent Trust Hub on Oct 5, 2026
Risk Level: SAFEREMOTE_CODE_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
- [REMOTE_CODE_EXECUTION]: The skill fetches and executes an installation script from the official Linkerd domain (https://run.linkerd.io/install). This represents the standard, documented method for installing the Linkerd CLI.
- [INDIRECT_PROMPT_INJECTION]: The skill provides templates for Kubernetes manifests and CLI commands that ingest user-defined parameters.
- Ingestion points: Placeholder values in SKILL.md for resource names, namespaces, and configuration weights.
- Boundary markers: None present.
- Capability inventory: Shell execution of kubectl and linkerd CLI tools.
- Sanitization: No validation logic is implemented for the interpolated values.
Audit Metadata