Data Models

Defines data contracts for KnowAI SSE.

PlanetContext

Planet context containing core theme and values mapping.

class PlanetContext(BaseModel):
    theme: str
    values_map: Dict[str, float]

Fields:

• theme (str): Core theme
• values_map (Dict[str, float]): Value weight mapping

Example:

context = PlanetContext(
    theme="Embodied Intelligence",
    values_map={"radical": 0.8, "ethics": 0.2}
)

SearchInstruction

Search instruction defining a single search task.

class SearchInstruction(BaseModel):
    channel: SearchChannel
    query: str
    time_range: str

Fields:

• channel (SearchChannel): Search channel (arxiv/web/rss)
• query (str): Search query
• time_range (str): Time range (latest/past_24h/past_week/past_month)

SearchChannel

Search channel enum.

class SearchChannel(str, Enum):
    ARXIV = "arxiv"
    WEB = "web"
    RSS = "rss"

SSEOutput

Expander output containing search instructions.

class SSEOutput(BaseModel):
    instructions: List[SearchInstruction]

Fields:

• instructions (List[SearchInstruction]): List of search instructions

Example:

result = await expander.expand(context)
for instruction in result.instructions:
    print(f"[{instruction.channel}] {instruction.query}")