import { Command } from 'commander';
import { ICommand } from '../lib/command-registry.js';
import { MergeConfiguration } from '../types.js';
/**
 * CLI command for retrieving complete documentation content by URI.
 *
 * This command complements `docs:search` by providing a way to fetch the full
 * content of a document after identifying it through search. The typical workflow:
 *
 * 1. User searches: `grepr docs:search "datadog" -f compact`
 *    → Gets preview with URI: doc://integrations/datadog/page.mdx
 *
 * 2. User retrieves: `grepr docs:get "doc://integrations/datadog/page.mdx"`
 *    → Gets complete markdown content
 *
 * Design decisions:
 *
 * - **No format options**: Documents are already in markdown format, so we output
 *   them as-is. This makes the command simple and perfect for piping to other tools
 *   (e.g., AI assistants, markdown processors).
 *
 * - **Option A retrieval**: Uses direct document lookup via `DocsSearch.getDocument()`
 *   which internally calls `listDocuments()` and `renderAllSections()`. This is more
 *   efficient than query-based retrieval and returns sections in document order
 *   rather than relevance order.
 *
 * - **Clean output**: No headers, colors, or formatting. Just the raw markdown
 *   content. This is intentional for scriptability and AI integration.
 *
 * Example usage:
 *   grepr docs:get "doc://tutorials/first-pipeline/page.mdx"
 *   grepr docs:get "doc://integrations/datadog/page.mdx" | some-ai-tool
 */
export declare class DocsGetCommand implements ICommand {
    getCommandName(): string;
    getCommandDescription(): string;
    addToProgram(program: Command, _mergeConfiguration: MergeConfiguration): void;
    /**
     * Executes the document retrieval command.
     *
     * This method:
     * 1. Opens the bundled documentation index without loading the embedding model
     * 2. Retrieves the full document content by URI
     * 3. Outputs raw markdown to stdout (no formatting, perfect for piping)
     *
     * Error handling: If the document is not found, DocsSearch.getDocument() throws
     * an error which is caught by the action handler and displayed to the user.
     *
     * @param uri - Document URI in format 'doc://path/to/file.mdx'
     */
    private execute;
}
//# sourceMappingURL=docs-get-command.d.ts.map