The Agent Skills Directory

[COMMAND_EXECUTION]: The skill uses a variety of system commands to perform diagnostics, including launchctl for service management, kubectl for Kubernetes cluster status, colima for container runtime status, and a custom joelclaw CLI. It also performs a reachability check using curl to the Anthropic API with a dummy authentication key.
[INDIRECT_PROMPT_INJECTION]: The skill's primary function is to ingest and display diagnostic logs and session transcripts, which represents an attack surface for indirect prompt injection if the logs contain malicious instructions.
Ingestion points: Log files (/tmp/joelclaw/gateway.log, /tmp/joelclaw/gateway.err) and session transcripts (~/.joelclaw/sessions/gateway/*.jsonl).
Boundary markers: None identified in the diagnostic procedures; logs are read directly using tail or cat.
Capability inventory: The skill has access to execute shell commands, perform Kubernetes operations (kubectl exec), and make network requests (curl).
Sanitization: No explicit sanitization or filtering of the log content is described before it is processed by the agent.

gateway-diagnose