Skip to content

Knowledge Chunker¶

Cluster: Uncategorised | Type: component | MCP Tools: None

Overview¶

Public API¶

`Passage`¶

Field	Type	Default
`chunk_id`	`str`	required
`text`	`str`	required
`domain`	`str`	required
`source_url`	`str`	required
`page_num`	`Optional[int]`	required
`timestamp_start`	`Optional[float]`	required
`timestamp_end`	`Optional[float]`	required
`frame_idx`	`Optional[int]`	required
`token_count`	`int`	required

Functions¶

`chunk_text(text: str, source_url: str, domain: str, chunk_size: int = 256, stride: int = 32, page_num: Optional[int] = None, timestamp_start: Optional[float] = None, timestamp_end: Optional[float] = None, frame_idx: Optional[int] = None, manifest_db: Optional['ManifestDB'] = None) -> list[Passage]`¶

Split text into overlapping token chunks with paragraph-boundary awareness.