# Generated by API Evangelist (build-phrasing.py). Our phrasing, not observed demand. overlay: 1.0.0 info: title: API Evangelist conversational phrasing for AI impact measurement Model API version: 1.0.0 extends: openapi/scope3-model-api-openapi.yml actions: - target: $.info update: x-apievangelist-phrasing: method: generated generated: '2026-10-01' generator: build-phrasing.py label: Generated by API Evangelist operations: 7 - target: $.paths['/model'].get update: x-apievangelist-phrasing: intent: List AI models effect: read questions: - Which AI models are available for impact measurement? - Can I list models from one model family only? instructions: - text: List all AI models. - text: List models in family {family}. slots: family: query.family method: generated generated: '2026-10-01' - target: $.paths['/model'].post update: x-apievangelist-phrasing: intent: Register a custom AI model effect: write questions: - How do I add my own fine-tuned model so its footprint can be estimated? - Can I record training energy and emissions for a custom model? instructions: - text: Create model {id}. slots: id: requestBody.id - text: Register model {id} named {name} with {total_params_billions} billion parameters. slots: id: requestBody.id name: requestBody.name total_params_billions: requestBody.total_params_billions - text: Add custom model {id} fine-tuned from {fine_tuned_from_model_id}. slots: id: requestBody.id fine_tuned_from_model_id: requestBody.fine_tuned_from_model_id method: generated generated: '2026-10-01' - target: $.paths['/model/{modelId}'].get update: x-apievangelist-phrasing: intent: Get an AI model's details effect: read questions: - What parameters and training footprint are recorded for a model? - Can I look up a global or custom model by ID? instructions: - text: Get model {modelId}. slots: modelId: path.modelId - text: Show the training emissions recorded for model {modelId}. slots: modelId: path.modelId method: generated generated: '2026-10-01' - target: $.paths['/model/{modelId}'].put update: x-apievangelist-phrasing: intent: Update an AI model effect: write questions: - Can I correct the parameter count or training energy on a model? - How do I change the estimated requests per day for a model? instructions: - text: Set training energy of model {modelId} to {training_usage_energy_kwh} kWh. slots: modelId: path.modelId training_usage_energy_kwh: requestBody.training_usage_energy_kwh - text: Update model {modelId} to {estimated_requests_per_day} requests per day. slots: modelId: path.modelId estimated_requests_per_day: requestBody.estimated_requests_per_day method: generated generated: '2026-10-01' - target: $.paths['/model/{modelId}'].delete update: x-apievangelist-phrasing: intent: Delete an AI model effect: destructive questions: - Can I remove a custom model I registered? - Is deleting a model permanent? instructions: - text: Delete model {modelId}. slots: modelId: path.modelId - text: Remove custom model {modelId}. slots: modelId: path.modelId method: generated generated: '2026-10-01' - target: $.paths['/model/{modelId}/alias'].put update: x-apievangelist-phrasing: intent: Add an alias to a model effect: write questions: - Can a model be known by more than one name? - How do I map an alternate model name to an existing model? instructions: - text: Add alias {alias} to model {modelId}. slots: alias: requestBody.alias modelId: path.modelId - text: Let {alias} also refer to model {modelId}. slots: alias: requestBody.alias modelId: path.modelId method: generated generated: '2026-10-01' - target: $.paths['/model/{modelId}/alias/{alias}'].delete update: x-apievangelist-phrasing: intent: Remove an alias from a model effect: destructive questions: - Can I drop an old alias from a model? - What happens to an alias I remove? instructions: - text: Remove alias {alias} from model {modelId}. slots: alias: path.alias modelId: path.modelId - text: Stop {alias} pointing to model {modelId}. slots: alias: path.alias modelId: path.modelId method: generated generated: '2026-10-01'