The Agent Skills Directory

[SAFE]: The skill focuses on read-only diagnostic operations for debugging wearable device integrations. It explicitly instructs the agent to avoid mutating application or device states unless requested by the user.
[INDIRECT_PROMPT_INJECTION]: The skill ingests data from external sources that could potentially contain malicious instructions intended to influence the agent's behavior.
Ingestion points: The agent reads live application logs, SDK events via wait_for_events, and error messages via get_errors.
Boundary markers: No specific boundary markers or instructions to ignore embedded prompt injections are present in the skill definition.
Capability inventory: The skill possesses the ability to connect to local debug servers and generate diagnostic files via export_diagnostic_bundle.
Sanitization: The instructions recommend generating a "redacted bundle" for diagnostics, which serves as a measure against unintentional data exposure.

live-debugging-mcp