Domain Sections — Python SDK

Domain sections attach business context to your traces, giving Maestro richer signal for evaluation. Each section is a dataclass you can pass as a keyword argument to send_task(), create_task_data(), or TraceBuilder.set_section().

How to Use

# As keyword argument
client.send_task(
    name="Handle ticket",
    description="Resolved a billing inquiry.",
    duration=3.2,
    customer={"customer_name": "Acme Corp", "customer_email": "[email protected]"},
    support={"issue_type": "billing", "priority": "high", "resolution": "Refund issued"},
)

# With TraceBuilder
trace = TraceBuilder("Handle ticket", "Resolved a billing inquiry.")
trace.set_section("customer", Customer(customer_name="Acme Corp"))
trace.set_section("support", Support(issue_type="billing", priority="high"))

You can pass either a dataclass instance or a plain dict (auto-coerced).

`Customer`

Context about the customer the agent is serving.

from infinium import Customer

Field	Type	Default	Description
`customer_name`	`str`	`None`	Customer’s name
`customer_email`	`str`	`None`	Customer’s email
`customer_phone`	`str`	`None`	Customer’s phone number
`customer_address`	`str`	`None`	Customer’s address
`client_company`	`str`	`None`	Company name
`client_industry`	`str`	`None`	Industry vertical

Example use case: Customer support agent, account management bot.

`Support`

Context for support/helpdesk interactions.

from infinium import Support

Field	Type	Default	Description
`call_id`	`str`	`None`	Ticket or call identifier
`issue_description`	`str`	`None`	Description of the issue
`issue_type`	`str`	`None`	Category (billing, technical, etc.)
`resolution`	`str`	`None`	How the issue was resolved
`priority`	`str`	`None`	Priority level (low, medium, high, critical)
`follow_up_required`	`str`	`None`	Whether follow-up is needed
`agent_notes`	`str`	`None`	Internal notes

Example use case: Ticket classifier, auto-resolution agent.

`Sales`

Context for sales pipeline activities.

from infinium import Sales

Field	Type	Default	Description
`lead_source`	`str`	`None`	Where the lead came from
`sales_stage`	`str`	`None`	Pipeline stage (prospecting, negotiation, etc.)
`deal_value`	`str`	`None`	Deal value
`conversion_rate`	`str`	`None`	Conversion rate
`sales_notes`	`str`	`None`	Sales notes

Example use case: Lead scoring agent, proposal generator.

`Marketing`

Context for marketing campaigns and activities.

from infinium import Marketing

Field	Type	Default	Description
`campaign_name`	`str`	`None`	Campaign identifier
`campaign_type`	`str`	`None`	Type (email, social, ads, etc.)
`target_audience`	`str`	`None`	Target demographic
`marketing_channel`	`str`	`None`	Distribution channel
`engagement_metrics`	`str`	`None`	Engagement data
`conversion_metrics`	`str`	`None`	Conversion data

Example use case: Content generator, audience segmentation agent.

`Content`

Context for content creation tasks.

from infinium import Content

Field	Type	Default	Description
`content_type`	`str`	`None`	Type (blog, email, social post, etc.)
`content_format`	`str`	`None`	Format (markdown, HTML, plain text)
`content_length`	`str`	`None`	Length or word count
`content_topic`	`str`	`None`	Topic or subject
`target_platform`	`str`	`None`	Publishing platform

Example use case: Blog writer, social media manager, newsletter generator.

`Research`

Context for research and analysis tasks.

from infinium import Research

Field	Type	Default	Description
`research_topic`	`str`	`None`	Topic being researched
`research_method`	`str`	`None`	Methodology used
`data_sources`	`str`	`None`	Sources consulted
`research_findings`	`str`	`None`	Summary of findings
`analysis_type`	`str`	`None`	Type of analysis
`sources_consulted`	`str`	`None`	Number/list of sources
`key_findings`	`str`	`None`	Key takeaways
`confidence_level`	`str`	`None`	Confidence in findings
`data_quality`	`str`	`None`	Quality assessment of data

Example use case: Market research agent, competitive analysis bot.

`Project`

Context for project management activities.

from infinium import Project

Field	Type	Default	Description
`project_name`	`str`	`None`	Project identifier
`project_phase`	`str`	`None`	Current phase
`deliverables`	`str`	`None`	Expected deliverables
`stakeholders`	`str`	`None`	Key stakeholders
`project_status`	`str`	`None`	Current status
`milestone_achieved`	`str`	`None`	Milestone completed
`task_priority`	`str`	`None`	Task priority
`resource_utilization`	`str`	`None`	Resource usage
`risk_assessment`	`str`	`None`	Risk level
`blockers`	`str`	`None`	Current blockers

Example use case: Sprint planning agent, status report generator.

`Development`

Context for software development tasks.

from infinium import Development

Field	Type	Default	Description
`programming_language`	`str`	`None`	Language used
`framework_used`	`str`	`None`	Framework or library
`bugs_found`	`str`	`None`	Number of bugs found
`bugs_fixed`	`str`	`None`	Number of bugs fixed
`test_coverage`	`str`	`None`	Test coverage percentage
`performance_metrics`	`str`	`None`	Performance data
`deployment_status`	`str`	`None`	Deployment state
`technical_debt`	`str`	`None`	Tech debt assessment
`files_modified`	`str`	`None`	Files changed
`tests_run`	`str`	`None`	Number of tests run
`tests_passed`	`str`	`None`	Number of tests passed

Example use case: Code review agent, test generation bot, CI/CD automation.

`Executive`

Context for executive and meeting activities.

from infinium import Executive

Field	Type	Default	Description
`meeting_type`	`str`	`None`	Type of meeting
`attendees_count`	`str`	`None`	Number of attendees
`agenda_items`	`str`	`None`	Meeting agenda
`action_items`	`str`	`None`	Action items generated
`calendar_conflicts`	`str`	`None`	Scheduling conflicts
`decisions_made`	`str`	`None`	Decisions reached

Example use case: Meeting summarizer, agenda planner, action item tracker.

`General`

Catch-all section for general-purpose context.

from infinium import General

Field	Type	Default	Description
`tools_used`	`str`	`None`	Tools the agent used
`agent_notes`	`str`	`None`	Free-form agent notes
`errors_encountered`	`str`	`None`	Errors hit during execution
`retries`	`str`	`None`	Number of retries
`warnings`	`str`	`None`	Warnings generated
`external_apis_called`	`str`	`None`	External APIs called
`data_sources_accessed`	`str`	`None`	Data sources used
`tags`	`str`	`None`	Comma-separated tags

Example use case: Any agent that doesn’t fit a specific domain.

`TimeTracking`

Simple start/end time tracking.

from infinium import TimeTracking

Field	Type	Default	Description
`start_time`	`str`	`None`	ISO 8601 start timestamp
`end_time`	`str`	`None`	ISO 8601 end timestamp

Multiple Sections

You can attach multiple domain sections to a single trace:

client.send_task(
    name="Research & report",
    description="Researched market trends and generated a report.",
    duration=12.5,
    research={"research_topic": "AI market trends", "sources_consulted": "5"},
    content={"content_type": "report", "content_length": "2500 words"},
    customer={"client_company": "Acme Corp", "client_industry": "Technology"},
)