The Agent Skills Directory

Indirect Prompt Injection Surface: The skill facilitates the ingestion of application logs and telemetry data using tools such as aspire logs and kubectl logs. This process introduces a surface where data originating from external sources could potentially include instructions for the agent. This is a standard consideration for observability tools and is typically addressed by platform-level guardrails.
Ingestion points: Data retrieved via aspire and kubectl logs as seen in SKILL.md and references/monitoring.md.
Capability inventory: Access to diagnostic commands and resource state information via the Aspire and Azure CLIs.
Boundary markers: No explicit delimiters are specified for log data isolation in the skill instructions.
Sanitization: The instructions do not define specific sanitization or filtering procedures for ingested log content.
Diagnostic Tool Routing: The skill provides a structured framework for choosing between local and cloud-based diagnostic tools. This routing logic is a functional component that ensures observability tasks are directed to the correct environment, such as using platform-specific diagnostics for deployed resources or kubectl for Kubernetes clusters.
Management of Telemetry Access: Documentation within the skill describes the use of tokens and API keys to secure access to the Aspire Dashboard. These elements are handled as dynamic parameters during local tool operation and are not hardcoded, which is consistent with secure development workflows for managing telemetry streams.

aspire-monitoring