/**
 * Document chunking — turning a document's extracted text into the enriched
 * chunks that get embedded and stored.
 *
 * Two passes, in this order:
 *  1. CHUNK — split on heading/paragraph boundaries (shared `@monoes/memory`
 *     chunker, with an inline fallback kept byte-identical to it).
 *  2. ENRICH — replace each chunk's leaf-heading prefix with a situating blurb
 *     (doc title + full heading path + summary) so the embedding model can tell
 *     two identically-titled sections of different documents apart.
 *
 * Split out of document-pipeline.ts. Pure string work: no fs, no network, no
 * store access — which is what makes chunk spans reproducible at citation time
 * (`chunkSpans`) without re-reading anything.
 *
 * @module v1/cli/knowledge/document-chunking
 */
import type { ChunkSpan } from './citation.js';
export interface TextChunk {
    chunkId: string;
    docId: string;
    text: string;
    startChar: number;
    endChar: number;
    chunkIndex: number;
}
export declare function chunkDocument(docId: string, text: string): Promise<TextChunk[]>;
/**
 * Replace each chunk's `§ heading` prefix with a richer situating blurb:
 *   § <doc title> · <full heading path>
 *   <doc summary for non-first chunks>
 *
 * First chunks that start with their own heading are left untouched (the
 * heading IS the context). The summary line is omitted for the first
 * chunk since it is adjacent to the summary text anyway.
 */
export declare function enrichChunks(chunks: TextChunk[], fullText: string, filePath: string): TextChunk[];
/** Chunk spans for a text, using the SAME chunker the ingest used — this is
 *  what makes a chunk index resolvable back to a character range without
 *  storing the text twice. */
export declare function chunkSpans(text: string): Promise<ChunkSpan[]>;
//# sourceMappingURL=document-chunking.d.ts.map