Doclea MCP

import { existsSync, readdirSync, readFileSync, statSync } from "node:fs"; import { basename, join, relative } from "node:path"; import { z } from "zod"; import type { SQLiteDatabase } from "@/database/sqlite"; import type { EmbeddingClient } from "@/embeddings/provider"; import { storeMemory } from "@/tools/memory"; import type { MemoryType } from "@/types"; import type { VectorStore } from "@/vectors/interface"; export const ImportInputSchema = z.object({ source: z.enum(["markdown", "adr"]).describe("Import source type"), path: z.string().describe("Path to import from (file or directory)"), projectPath: z .string() .optional() .describe("Project path. Defaults to current directory."), recursive: z.boolean().default(true).describe("Recursively scan directories"), dryRun: z.boolean().default(false).describe("Preview without storing"), }); export type ImportInput = z.infer<typeof ImportInputSchema>; export interface ImportResult { imported: number; skipped: number; files: string[]; errors: string[]; } export async function importContent( input: ImportInput, db: SQLiteDatabase, vectors: VectorStore, embeddings: EmbeddingClient, ): Promise<ImportResult> { const projectPath = input.projectPath ?? process.cwd(); const targetPath = join(projectPath, input.path); const result: ImportResult = { imported: 0, skipped: 0, files: [], errors: [], }; if (!existsSync(targetPath)) { result.errors.push(`Path not found: ${input.path}`); return result; } const stat = statSync(targetPath); const files = stat.isDirectory() ? findMarkdownFiles(targetPath, input.recursive) : [targetPath]; for (const file of files) { try { const content = readFileSync(file, "utf-8"); const relPath = relative(projectPath, file); if (input.source === "adr") { const adr = parseADR(content, relPath); if (adr) { if (!input.dryRun) { await storeMemory( { type: "decision" as MemoryType, title: adr.title, content: adr.content, summary: adr.summary, importance: 0.9, tags: ["adr", "imported", ...adr.tags], relatedFiles: [relPath], experts: [], }, db, vectors, embeddings, ); } result.imported++; result.files.push(relPath); } else { result.skipped++; } } else { // Generic markdown import const doc = parseMarkdown(content, relPath); if (doc.content.length > 100) { if (!input.dryRun) { await storeMemory( { type: "note" as MemoryType, title: doc.title, content: doc.content.slice(0, 10000), importance: 0.5, tags: ["markdown", "imported"], relatedFiles: [relPath], experts: [], }, db, vectors, embeddings, ); } result.imported++; result.files.push(relPath); } else { result.skipped++; } } } catch (error) { result.errors.push( `Failed to process ${file}: ${error instanceof Error ? error.message : "Unknown error"}`, ); } } return result; } function findMarkdownFiles(dir: string, recursive: boolean): string[] { const results: string[] = []; const entries = readdirSync(dir); for (const entry of entries) { const fullPath = join(dir, entry); const stat = statSync(fullPath); if (stat.isDirectory() && recursive) { results.push(...findMarkdownFiles(fullPath, recursive)); } else if (entry.match(/\.(md|mdx)$/i)) { results.push(fullPath); } } return results; } interface ParsedADR { title: string; content: string; summary: string; status: string; tags: string[]; } function parseADR(content: string, filename: string): ParsedADR | null { // Try to parse ADR format (MADR or similar) const titleMatch = content.match(/^#\s+(.+)$/m); const statusMatch = content.match(/##\s*Status\s*\n+(\w+)/i); const contextMatch = content.match(/##\s*Context\s*\n+([\s\S]*?)(?=\n##|$)/i); const decisionMatch = content.match( /##\s*Decision\s*\n+([\s\S]*?)(?=\n##|$)/i, ); if (!titleMatch) { // Not a valid ADR return null; } const title = titleMatch[1].replace(/^ADR[-\s]*\d*:?\s*/i, "").trim(); const status = statusMatch?.[1] ?? "unknown"; const context = contextMatch?.[1]?.trim() ?? ""; const decision = decisionMatch?.[1]?.trim() ?? ""; const summary = context.split("\n")[0]?.slice(0, 200) ?? ""; // Extract tags from content const tags: string[] = []; if (status.toLowerCase() === "accepted") tags.push("accepted"); if (status.toLowerCase() === "deprecated") tags.push("deprecated"); if (status.toLowerCase() === "superseded") tags.push("superseded"); return { title, content: `## Context\n${context}\n\n## Decision\n${decision}`, summary, status, tags, }; } interface ParsedMarkdown { title: string; content: string; } function parseMarkdown(content: string, filename: string): ParsedMarkdown { // Extract title from first heading or filename const titleMatch = content.match(/^#\s+(.+)$/m); const title = titleMatch?.[1] ?? basename(filename, ".md").replace(/[-_]/g, " "); return { title, content, }; }