deploy-k8s-explorer
Pass
Audited by Gen Agent Trust Hub on Aug 18, 2026
Risk Level: SAFECOMMAND_EXECUTIONEXTERNAL_DOWNLOADSPROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill makes extensive use of local command execution to interact with the user's environment. This includes standard Kubernetes management tools (
kubectl) and the vendor's own CLI (observe). These commands are used for cluster discovery, resource creation, and verification of pod statuses. - [EXTERNAL_DOWNLOADS]: The skill instructs the user to install the OpenTelemetry Operator via Helm using the official
open-telemetryrepository. This is a well-known and trusted source within the cloud-native ecosystem. Additionally, it references documentation and configuration hosted on the author's official domain (docs.observeinc.com). - [PROMPT_INJECTION]: The skill exhibits high security awareness regarding indirect prompt injection. It acknowledges that data ingested from Kubernetes clusters (logs, pod metadata, event messages) is untrusted. To mitigate this, it implements a robust 'wrap' helper—a shell function that encapsulates untrusted command output within nonce-delimited tags and sanitizes ANSI escape sequences. This prevents malicious data from being interpreted as instructions by the agent.
- [SAFE]: No hardcoded credentials, malicious persistence mechanisms, or obfuscated payloads were detected. The workflow adheres to least-privilege principles by asking the user for confirmation at critical steps (authentication, resource creation) and using environment variables for sensitive tokens rather than injecting them directly into the agent's context.
Audit Metadata