| name | generate-node |
| description | Gemini-native ONEX node generation — generates production-ready nodes with whole-project grounding |
| capability | skill.generate_node |
| node_name | node_skill_generate_node_orchestrator |
| dot_color | 39 |
Generate Node Skill (Gemini Edition)
Fully automated ONEX node generation that leverages Gemini's massive context window to ensure new nodes are perfectly integrated with existing patterns, contracts, and models.
Workflow
- Inference: Gemini analyzes the generation prompt against the entire codebase to infer the optimal contract, input/output models, and node type.
- Template Expansion: Generates the infrastructure boilerplate using existing Jinja2 templates.
- Logic Generation: Gemini generates the complete business logic, automatically identifying and reusing existing internal APIs.
- Validation: Runs generated tests and quality checks to ensure ONEX v2.0 compliance.
Gemini Advantages
- Perfect Pattern Matching: New nodes automatically adopt the specific architectural style of the surrounding modules without manual guidance.
- Superior Contract Accuracy: Gemini correctly identifies the necessary Kafka topics and Pydantic models by seeing the whole organization's registry.
- High-Fidelity Tests: Generates comprehensive unit tests that cover realistic edge cases derived from the project's actual usage patterns.
Arguments
PROMPT: Natural language description of the node.
--output-dir: Output directory for generated files.
--node-type: Hint for the node type (effect, compute, reducer, orchestrator).
--enable-quorum: Use multi-model validation for the generated code.