What problem does it solve? Working with Genkit projects requires knowing flow input schemas, managing background runtimes, and debugging failed executions, which is error-prone when done blindly through manual commands. ## Core Features & Use Cases - Flow Discovery and Execution: List all defined Genkit flows with their input schemas via list_flows, then run them safely with run_flow using schema-validated JSON input. - Trace-Based Debugging: Retrieve detailed execution traces with get_trace using a traceId to diagnose errors and analyze performance instead of guessing from surface errors. - Runtime Lifecycle Control: Start, restart, or stop the Genkit development server with start_runtime, restart_runtime, and kill_runtime when connections fail. - Documentation Lookup: Query official Genkit guides and docs with get_usage_guide, search_genkit_docs, and read_genkit_docs to avoid outdated API usage. - Use Case: When a flow execution returns an unexpected result, extract the traceId from the response and call get_trace to pinpoint the failing step before changing any code. ## Quick Start Ask the agent to list all available Genkit flows and their input schemas, then run a specific flow with a JSON payload matching that schema.
