mounted_skill

Overview

This skill reports which CCB providers are mounted for the current project. It checks that a provider has an active session and that the background daemon is online, then outputs the result in JSON for easy consumption by scripts or tooling.

How this skill works

When executed, the skill queries the local CCB runtime to enumerate providers tied to the current project. For each provider it evaluates two conditions: whether a session exists and whether the daemon is reachable; providers meeting both conditions are reported as mounted. The final output is a compact JSON object listing mounted providers and relevant status flags.

When to use it

• Verify which AI providers are available before running multi-agent workflows
• Integrate with CI/CD checks to ensure required providers are online
• Trigger conditional automation based on mounted provider presence
• Debug connectivity or session issues between the CLI and provider daemons
• Feed provider status into dashboards or monitoring pipelines

Best practices

• Run the check from the project root so provider contexts resolve correctly
• Use the JSON output in scripts to programmatically gate downstream tasks
• Run frequently in ephemeral environments to detect transient daemon failures
• Combine with provider logs when a provider appears unmounted to diagnose issues
• Treat the mounted check as a fast preflight rather than a deep health check

Example use cases

• A build script aborts when a required LLM provider is not mounted to avoid unpredictable runs
• A monitoring job polls mounted providers and alerts when a daemon goes offline
• An orchestration tool selects alternate providers when the primary provider is unmounted
• A developer script lists mounted providers prior to launching interactive sessions