The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill instructs the agent to process external, potentially untrusted data that could contain malicious instructions designed to influence the agent's behavior.
Ingestion points: Data is ingested via rpk debug bundle (which collects system logs and crash reports) and curl (which fetches Redpanda Admin API responses).
Boundary markers: The skill does not define specific delimiters or instructions to ignore embedded commands within the logs or API data provided to the agent.
Capability inventory: The skill uses high-privilege diagnostic tools including rpk, curl, kubectl, and journalctl, as well as file management commands like unzip and cat.
Sanitization: There is no mention of sanitizing or filtering the content of logs or API responses before they are processed by the agent.
[COMMAND_EXECUTION]: The skill heavily relies on the execution of CLI tools for system and cluster diagnostics.
Evidence: Extensive use of rpk, curl, kubectl, journalctl, jq, and unzip across all documentation files.
Context: These commands are directed at local endpoints (localhost:9644) or vendor-controlled resources (e.g., Redpanda support upload URLs on AWS S3). The diagnostic commands are standard for the management of the Redpanda platform and align with the skill's primary purpose.

streaming-debugging