The Agent Skills Directory

[COMMAND_EXECUTION]: The skill instructs the agent to run the flutter-mcp-toolkit doctor --json command to verify the local environment and troubleshoot connectivity issues.
[INDIRECT_PROMPT_INJECTION]: The toolkit provides a surface for indirect prompt injection as it ingests untrusted data from running Flutter applications (UI widgets, logs, and errors) into the agent's context.
Ingestion points: App state and logs are retrieved via tools such as get_view_details, semantic_snapshot, get_app_errors, and get_recent_logs (SKILL.md).
Boundary markers: No explicit boundary markers or instruction-ignoring warnings are mentioned in this routing guide.
Capability inventory: The toolkit includes powerful capabilities such as tap_widget, enter_text, evaluate_dart_expression, and hot_restart_flutter (SKILL.md).
Sanitization: No sanitization or filtering logic is described for the ingested application data.
[DYNAMIC_EXECUTION]: References the evaluate_dart_expression tool within the debug category, which allows the execution of arbitrary Dart code within the context of the connected Flutter application.

flutter-mcp-toolkit-guide