prompt = client.prompts.create(options)
# Async
prompt = await client.prompts.acreate(options)
Parameters
Pass a CreatePromptOptions object:
| Parameter | Type | Required | Description |
|---|
slug | str | Yes | URL-friendly identifier |
name | str | Yes | Display name for the prompt |
messages | list[PromptMessage] | Yes | Array of messages |
model | str | Yes | Default model (e.g., gpt-4o) |
provider | LLMProvider | Yes | Provider ("openai", "anthropic", "google") |
description | str | No | Description of the prompt |
temperature | float | No | Default temperature |
max_output_tokens | int | No | Default max output tokens |
top_p | float | No | Default top-p value |
stop_sequences | list[str] | No | Default stop sequences |
PromptMessage
| Field | Type | Required | Description |
|---|
role | str | Yes | system, developer, user, or assistant |
content | str | Yes | Message content (supports {{variables}}) |
Response
Returns the created Prompt object.
Examples
Basic Usage
from tracia import CreatePromptOptions, PromptMessage
prompt = client.prompts.create(CreatePromptOptions(
slug="welcome-email",
name="Welcome Email",
model="gpt-4o",
provider="openai",
messages=[
PromptMessage(role="system", content="You are a helpful assistant."),
PromptMessage(role="user", content="Write a welcome email for {{name}}."),
],
))
print(prompt.slug) # "welcome-email"
With Description and Options
prompt = client.prompts.create(CreatePromptOptions(
slug="onboarding-welcome",
name="Welcome Email",
description="Sent to new users after signup",
model="gpt-4o",
provider="openai",
temperature=0.7,
messages=[
PromptMessage(role="system", content="You write professional emails."),
PromptMessage(role="user", content="Write a welcome email for {{name}} at {{company}}."),
],
))
Multi-Turn Conversation
prompt = client.prompts.create(CreatePromptOptions(
slug="code-review",
name="Code Review Assistant",
model="claude-sonnet-4-20250514",
provider="anthropic",
messages=[
PromptMessage(
role="system",
content="You are a senior software engineer conducting code reviews.",
),
PromptMessage(
role="user",
content="Please review this {{language}} code:\n\n{{code}}",
),
PromptMessage(
role="assistant",
content="I'll review this code for correctness, performance, and best practices.",
),
PromptMessage(
role="user",
content="Focus especially on {{focus_area}}.",
),
],
))
Slugs must be lowercase with hyphens only (e.g., welcome-email).
Error Handling
from tracia import TraciaError, TraciaErrorCode
try:
prompt = client.prompts.create(CreatePromptOptions(
slug="welcome-email",
name="Welcome Email",
model="gpt-4o",
provider="openai",
messages=[...],
))
except TraciaError as error:
if error.code == TraciaErrorCode.CONFLICT:
print("A prompt with this slug already exists")
elif error.code == TraciaErrorCode.INVALID_REQUEST:
print(f"Invalid request: {error.message}")
