diff --git a/apps/api/src/common/cbr.ts b/apps/api/src/common/cbr.ts
index cbeacad..76f44d2 100644
--- a/apps/api/src/common/cbr.ts
+++ b/apps/api/src/common/cbr.ts
@@ -1,4 +1,4 @@
-import { mkdtempSync, readFileSync, rmSync } from "node:fs";
+import { mkdirSync, mkdtempSync, readFileSync, rmSync } from "node:fs";
 import { extname, join } from "node:path";
 import { createExtractorFromFile } from "node-unrar-js";
 import { COMIC_IMAGE_EXTENSIONS, MAX_COMIC_ARCHIVE_ENTRIES } from "./cbz.js";
@@ -46,6 +46,7 @@ export async function readCbrPage(
     throw new Error("CBR page not found");
   }
 
+  mkdirSync(storageDir, { recursive: true });
   const tempDir = mkdtempSync(join(storageDir, "cbr-page-"));
   const safeName = `page${extname(page.entryName).toLowerCase() || ".jpg"}`;
   try {
diff --git a/apps/api/src/metadata/adapters/bnf.provider.ts b/apps/api/src/metadata/adapters/bnf.provider.ts
index 575d20d..bf35efa 100644
--- a/apps/api/src/metadata/adapters/bnf.provider.ts
+++ b/apps/api/src/metadata/adapters/bnf.provider.ts
@@ -2,6 +2,8 @@ import { Injectable } from "@nestjs/common";
 import { XMLParser } from "fast-xml-parser";
 import { MetadataLookup, MetadataMatch, MetadataProvider, MetadataProviderConfig, MetadataSearchQuery } from "../metadata.types.js";
 import { toIsbn13 } from "../use-cases/extract-identifiers.js";
+import { normalizePublishedDate } from "../use-cases/normalize-published-date.js";
+import { providerFetch, providerHttpError } from "./provider-fetch.js";
 
 const parser = new XMLParser({ ignoreAttributes: false, attributeNamePrefix: "@_", removeNSPrefix: true });
 
@@ -36,8 +38,8 @@ export class BnfProvider implements MetadataProvider {
     url.searchParams.set("query", query);
     url.searchParams.set("maximumRecords", String(maximumRecords));
 
-    const response = await fetch(url, { signal: AbortSignal.timeout(5000) });
-    if (!response.ok) return [];
+    const response = await providerFetch(this.id, url, { timeoutMs: 5000 });
+    if (!response.ok) throw await providerHttpError(this.id, response, `BnF HTTP ${response.status}`);
     const parsed = parser.parse(await response.text());
     const records = asArray(parsed?.searchRetrieveResponse?.records?.record)
       .map((entry) => (entry.recordData as Record<string, unknown> | undefined)?.record)
@@ -53,7 +55,7 @@ export class BnfProvider implements MetadataProvider {
           isbn: bestIsbn(fields, expectedIsbn13),
           language: subfield(fields, "101", "a"),
           publisher: subfield(fields, "210", "c") ?? subfield(fields, "214", "c"),
-          publishedDate: cleanDate(subfield(fields, "210", "d") ?? subfield(fields, "214", "d"))
+          publishedDate: normalizePublishedDate(cleanDate(subfield(fields, "210", "d") ?? subfield(fields, "214", "d")))
         };
       })
       .sort((left, right) => Number(Boolean(right.isbn)) - Number(Boolean(left.isbn)));
diff --git a/apps/api/src/metadata/adapters/comic-vine.provider.ts b/apps/api/src/metadata/adapters/comic-vine.provider.ts
new file mode 100644
index 0000000..bbbce6f
--- /dev/null
+++ b/apps/api/src/metadata/adapters/comic-vine.provider.ts
@@ -0,0 +1,160 @@
+import { Injectable } from "@nestjs/common";
+import { MetadataLookup, MetadataMatch, MetadataProvider, MetadataProviderConfig, MetadataSearchQuery } from "../metadata.types.js";
+import { normalizePublishedDate } from "../use-cases/normalize-published-date.js";
+import { providerFetch } from "./provider-fetch.js";
+
+@Injectable()
+export class ComicVineProvider implements MetadataProvider {
+  readonly id = "comicvine" as const;
+
+  async lookup(lookup: MetadataLookup, config: MetadataProviderConfig): Promise<MetadataMatch | null> {
+    const matches = await this.searchByMetadata({ title: lookup.title, author: lookup.author, year: lookup.local.year }, config);
+    return matches[0] ?? null;
+  }
+
+  async searchByMetadata(query: MetadataSearchQuery, config: MetadataProviderConfig): Promise<MetadataMatch[]> {
+    assertApiKey(config);
+    const title = cleanComicTitle(query.title);
+    const relaxed = title.replace(/\b\d{1,3}\b/g, " ").replace(/\s+/g, " ").trim();
+    const matches = [
+      ...(await searchComicVine("volume", title, config)),
+      ...(await searchComicVine("issue", title, config)),
+      ...(relaxed && relaxed !== title ? await searchComicVine("volume", relaxed, config) : [])
+    ];
+    return rankMatches(query.title, dedupe(matches));
+  }
+}
+
+export class ComicVineProviderError extends Error {
+  constructor(
+    readonly code: "missing-key" | "invalid-key" | "rate-limit" | "http",
+    readonly status: number,
+    message: string
+  ) {
+    super(message);
+    this.name = "ComicVineProviderError";
+  }
+}
+
+async function searchComicVine(resource: "volume" | "issue", title: string, config: MetadataProviderConfig): Promise<MetadataMatch[]> {
+  const url = new URL("https://comicvine.gamespot.com/api/search/");
+  url.searchParams.set("api_key", config.apiKey!);
+  url.searchParams.set("format", "json");
+  url.searchParams.set("resources", resource);
+  url.searchParams.set("query", title);
+  url.searchParams.set("limit", "10");
+  url.searchParams.set(
+    "field_list",
+    resource === "volume" ? "id,name,description,image,start_year,publisher" : "id,name,description,image,cover_date,store_date,volume"
+  );
+  const response = await providerFetch("comicvine", url, {
+    headers: { "User-Agent": "ReadaBook/0.1 self-hosted metadata provider (Comic Vine; non-commercial)" },
+    timeoutMs: 6000
+  });
+  const data = (await parseComicVineResponse(response)) as { results?: Array<Record<string, unknown>> };
+  return (data.results ?? []).map((entry) => comicVineToMatch(resource, entry));
+}
+
+function assertApiKey(config: MetadataProviderConfig): void {
+  if (!config.apiKey?.trim()) {
+    throw new ComicVineProviderError("missing-key", 0, "Comic Vine API key is required");
+  }
+}
+
+async function parseComicVineResponse(response: Response): Promise<unknown> {
+  const data = (await response.json().catch(() => ({}))) as { status_code?: number; error?: string };
+  if (response.status === 429) throw new ComicVineProviderError("rate-limit", response.status, data.error ?? "Comic Vine rate limit");
+  if (response.status === 401 || response.status === 403) throw new ComicVineProviderError("invalid-key", response.status, data.error ?? "Comic Vine API key rejected");
+  if (!response.ok) throw new ComicVineProviderError("http", response.status, data.error ?? `Comic Vine HTTP ${response.status}`);
+  if (data.status_code && data.status_code !== 1) {
+    const code = data.status_code === 100 || data.status_code === 101 ? "invalid-key" : "http";
+    throw new ComicVineProviderError(code, 200, data.error ?? `Comic Vine status ${data.status_code}`);
+  }
+  return data;
+}
+
+function comicVineToMatch(resource: "volume" | "issue", entry: Record<string, unknown>): MetadataMatch & { comicVineRank?: number } {
+  const volume = objectValue(entry.volume);
+  const title = resource === "issue" ? [stringValue(volume.name), stringValue(entry.name)].filter(Boolean).join(" ") : stringValue(entry.name);
+  return {
+    title: title || undefined,
+    description: cleanHtml(stringValue(entry.description)),
+    publisher: stringValue(objectValue(entry.publisher).name),
+    publishedDate: normalizePublishedDate(resource === "volume" ? stringValue(entry.start_year) : yearFromDate(stringValue(entry.cover_date) ?? stringValue(entry.store_date))),
+    coverUrl: imageUrl(entry.image),
+    sourceId: stringValue(entry.id)
+  };
+}
+
+function cleanComicTitle(value: string): string {
+  return value
+    .replace(/\.[A-Za-z0-9]{2,5}$/g, " ")
+    .replace(/[._]+/g, " ")
+    .replace(/\b(FRENCH|TRUEFRENCH|MULTI|CBZ|CBR|EPUB|PDF|eBook|ebook|scan|digital)\b/gi, " ")
+    .replace(/\([^)]*\)/g, " ")
+    .replace(/\b(e?bdz|Paprika\+?|emuleCenter(?:\.|\s+)net)\b/gi, " ")
+    .replace(/\bT(?:ome)?\s*0?(\d{1,3})\b/gi, " $1 ")
+    .replace(/[+]+/g, " ")
+    .replace(/\s+-\s+/g, " ")
+    .replace(/\s*-\s*$/g, " ")
+    .replace(/\s+/g, " ")
+    .trim();
+}
+
+function rankMatches(originalTitle: string, matches: Array<MetadataMatch & { comicVineRank?: number }>): MetadataMatch[] {
+  return [...matches]
+    .map((match) => ({ ...match, comicVineRank: comicRank(originalTitle, match) }))
+    .sort((left, right) => (right.comicVineRank ?? 0) - (left.comicVineRank ?? 0))
+    .map(({ comicVineRank: _rank, ...match }) => match);
+}
+
+function comicRank(originalTitle: string, match: MetadataMatch): number {
+  let rank = tokenOverlap(cleanComicTitle(originalTitle), match.title ?? "") * 10;
+  if (match.coverUrl) rank += 1;
+  if (match.description) rank += 1;
+  return rank;
+}
+
+function dedupe(matches: MetadataMatch[]): MetadataMatch[] {
+  const seen = new Set<string>();
+  return matches.filter((match) => {
+    const key = [match.sourceId, match.title].filter(Boolean).join("|");
+    if (seen.has(key)) return false;
+    seen.add(key);
+    return true;
+  });
+}
+
+function cleanHtml(value: string | null): string | null {
+  if (!value) return null;
+  return value.replace(/<[^>]*>/g, " ").replace(/&nbsp;/g, " ").replace(/&amp;/g, "&").replace(/\s+/g, " ").trim() || null;
+}
+
+function imageUrl(value: unknown): string | null {
+  const image = objectValue(value);
+  return stringValue(image.original_url) ?? stringValue(image.super_url) ?? stringValue(image.medium_url) ?? stringValue(image.small_url);
+}
+
+function yearFromDate(value: string | null): string | null {
+  return value?.match(/\b(1[5-9]\d{2}|20\d{2})\b/)?.[1] ?? null;
+}
+
+function tokenOverlap(left: string, right: string): number {
+  const leftTokens = new Set(normalizeTokens(left));
+  const rightTokens = new Set(normalizeTokens(right));
+  if (!leftTokens.size || !rightTokens.size) return 0;
+  return [...leftTokens].filter((token) => rightTokens.has(token)).length / leftTokens.size;
+}
+
+function normalizeTokens(value: string): string[] {
+  return value.normalize("NFD").replace(/[\u0300-\u036f]/g, "").toLowerCase().replace(/[^a-z0-9]+/g, " ").split(" ").filter(Boolean);
+}
+
+function objectValue(value: unknown): Record<string, unknown> {
+  return value && typeof value === "object" && !Array.isArray(value) ? (value as Record<string, unknown>) : {};
+}
+
+function stringValue(value: unknown): string | null {
+  if (typeof value === "number") return String(value);
+  return typeof value === "string" && value.trim() ? value.trim() : null;
+}
diff --git a/apps/api/src/metadata/adapters/google-books.provider.ts b/apps/api/src/metadata/adapters/google-books.provider.ts
index 793a1d8..1fa6d74 100644
--- a/apps/api/src/metadata/adapters/google-books.provider.ts
+++ b/apps/api/src/metadata/adapters/google-books.provider.ts
@@ -1,6 +1,8 @@
 import { Injectable } from "@nestjs/common";
 import { MetadataLookup, MetadataMatch, MetadataProvider, MetadataProviderConfig, MetadataSearchQuery } from "../metadata.types.js";
 import { toIsbn13 } from "../use-cases/extract-identifiers.js";
+import { normalizePublishedDate } from "../use-cases/normalize-published-date.js";
+import { providerFetch } from "./provider-fetch.js";
 
 @Injectable()
 export class GoogleBooksProvider implements MetadataProvider {
@@ -12,54 +14,157 @@ export class GoogleBooksProvider implements MetadataProvider {
       const matches = await this.searchByMetadata({ title: lookup.title, author: lookup.author, year: lookup.local.year }, config);
       return matches[0] ?? null;
     }
-    const query = `isbn:${isbn}`;
-    const url = new URL("https://www.googleapis.com/books/v1/volumes");
-    url.searchParams.set("q", query);
-    url.searchParams.set("maxResults", "1");
-    url.searchParams.set("printType", "books");
-    if (config.apiKey) url.searchParams.set("key", config.apiKey);
-
-    const response = await fetch(url, { signal: AbortSignal.timeout(4000) });
-    if (!response.ok) return null;
-    const data = (await response.json()) as { items?: Array<{ volumeInfo?: Record<string, unknown> }> };
-    const info = data.items?.[0]?.volumeInfo;
-    if (!info) return null;
-    return {
-      title: stringValue(info.title) ?? undefined,
-      author: arrayJoin(info.authors),
-      description: stringValue(info.description),
-      language: stringValue(info.language),
-      publisher: stringValue(info.publisher),
-      publishedDate: stringValue(info.publishedDate),
-      isbn: isbnFromIndustryIdentifiers(info.industryIdentifiers, lookup.identifiers.isbn13)
-    };
+    const matches = await this.searchVolumes(`isbn:${isbn}`, config, lookup.title, lookup.author, lookup.identifiers.isbn13);
+    return rankMatches(lookup.title, matches)[0] ?? null;
   }
 
   async searchByMetadata(query: MetadataSearchQuery, config: MetadataProviderConfig): Promise<MetadataMatch[]> {
+    const attempts = googleBookQueries(query);
+    const matches: MetadataMatch[] = [];
+    const seen = new Set<string>();
+    for (const attempt of attempts) {
+      for (const match of await this.searchVolumes(attempt, config, query.title, query.author, query.isbn ? toIsbn13(query.isbn) : null)) {
+        const key = [match.sourceId, match.isbn, match.title, match.author].filter(Boolean).join("|");
+        if (seen.has(key)) continue;
+        seen.add(key);
+        matches.push(match);
+      }
+    }
+    return rankMatches(query.title, matches);
+  }
+
+  private async searchVolumes(
+    googleQuery: string,
+    config: MetadataProviderConfig,
+    originalTitle: string,
+    originalAuthor: string | null,
+    expectedIsbn13: string | null
+  ): Promise<MetadataMatch[]> {
     const url = new URL("https://www.googleapis.com/books/v1/volumes");
-    url.searchParams.set("q", `intitle:${query.title}${query.author ? `+inauthor:${query.author}` : ""}`);
-    url.searchParams.set("maxResults", "5");
+    url.searchParams.set("q", googleQuery);
+    url.searchParams.set("maxResults", "10");
     url.searchParams.set("printType", "books");
     if (config.apiKey) url.searchParams.set("key", config.apiKey);
 
-    const response = await fetch(url, { signal: AbortSignal.timeout(4000) });
-    if (!response.ok) return [];
-    const data = (await response.json()) as { items?: Array<{ volumeInfo?: Record<string, unknown> }> };
+    const response = await providerFetch(this.id, url, { timeoutMs: 4000 });
+    const data = (await parseGoogleResponse(response)) as { items?: Array<{ id?: string; volumeInfo?: Record<string, unknown> }> };
     return (data.items ?? [])
-      .map((item) => item.volumeInfo)
-      .filter((info): info is Record<string, unknown> => Boolean(info))
-      .map((info) => ({
+      .map((item) => ({ sourceId: item.id, info: item.volumeInfo }))
+      .filter((item): item is { sourceId: string | undefined; info: Record<string, unknown> } => Boolean(item.info))
+      .map(({ sourceId, info }) => ({
         title: stringValue(info.title) ?? undefined,
         author: arrayJoin(info.authors),
         description: stringValue(info.description),
         language: stringValue(info.language),
         publisher: stringValue(info.publisher),
-        publishedDate: stringValue(info.publishedDate),
-        isbn: isbnFromIndustryIdentifiers(info.industryIdentifiers, query.isbn ? toIsbn13(query.isbn) : null)
+        publishedDate: normalizePublishedDate(stringValue(info.publishedDate)),
+        isbn: isbnFromIndustryIdentifiers(info.industryIdentifiers, expectedIsbn13),
+        coverUrl: coverUrl(info.imageLinks),
+        sourceId,
+        identifiers: { candidates: isbnCandidates(info.industryIdentifiers) },
+        googleRank: googleRank(originalTitle, originalAuthor, info)
       }));
   }
 }
 
+export class GoogleBooksProviderError extends Error {
+  constructor(
+    readonly code: "quota" | "auth" | "http",
+    readonly status: number,
+    message: string
+  ) {
+    super(message);
+    this.name = "GoogleBooksProviderError";
+  }
+}
+
+async function parseGoogleResponse(response: Response): Promise<unknown> {
+  const data = (await response.json().catch(() => ({}))) as { error?: { message?: string; status?: string } };
+  if (response.ok) return data;
+  const message = data.error?.message ?? `Google Books HTTP ${response.status}`;
+  if (response.status === 429) throw new GoogleBooksProviderError("quota", response.status, message);
+  if (response.status === 401 || response.status === 403) throw new GoogleBooksProviderError("auth", response.status, message);
+  throw new GoogleBooksProviderError("http", response.status, message);
+}
+
+function googleBookQueries(query: MetadataSearchQuery): string[] {
+  const cleaned = cleanGoogleBooksTitle(query.title);
+  const relaxed = relaxSeriesTitle(cleaned);
+  return [
+    query.isbn ? `isbn:${query.isbn}` : null,
+    googleTitleQuery(cleaned, query.author, true),
+    googleTitleQuery(cleaned, query.author, false),
+    relaxed !== cleaned ? googleTitleQuery(relaxed, query.author, true) : null,
+    relaxed !== cleaned ? googleTitleQuery(relaxed, null, false) : null,
+    googleTitleQuery(cleaned, null, false)
+  ].filter((value, index, values): value is string => Boolean(value) && values.indexOf(value) === index);
+}
+
+function googleTitleQuery(title: string, author: string | null, quoted: boolean): string {
+  const titlePart = quoted ? `intitle:"${title.replace(/"/g, " ")}"` : `intitle:${title}`;
+  return author ? `${titlePart}+inauthor:${author}` : titlePart;
+}
+
+function cleanGoogleBooksTitle(value: string): string {
+  return value
+    .replace(/\.[A-Za-z0-9]{2,5}$/g, " ")
+    .replace(/[._]+/g, " ")
+    .replace(/\b(FRENCH|TRUEFRENCH|MULTI|CBZ|CBR|EPUB|PDF|eBook|ebook|scan|digital)\b/gi, " ")
+    .replace(/\b(e?bdz|Paprika\+?|emuleCenter\.net)\b/gi, " ")
+    .replace(/\bT(?:ome)?\s*0?(\d{1,3})\b/gi, " $1 ")
+    .replace(/\bVol(?:ume)?\.?\s*0?(\d{1,3})\b/gi, " $1 ")
+    .replace(/[+]+/g, " ")
+    .replace(/\s+-\s+/g, " ")
+    .replace(/\s*-\s*$/g, " ")
+    .replace(/\s+/g, " ")
+    .trim();
+}
+
+function relaxSeriesTitle(value: string): string {
+  return value.replace(/\b\d{1,3}\b/g, " ").replace(/\s+/g, " ").trim();
+}
+
+function rankMatches(originalTitle: string, matches: Array<MetadataMatch & { googleRank?: number }>): MetadataMatch[] {
+  return [...matches]
+    .sort((left, right) => (right.googleRank ?? 0) - (left.googleRank ?? 0))
+    .map(({ googleRank: _googleRank, ...match }) => match);
+}
+
+function googleRank(originalTitle: string, originalAuthor: string | null, info: Record<string, unknown>): number {
+  const expectedVolume = volumeNumber(originalTitle);
+  const candidateTitle = [stringValue(info.title), stringValue(info.subtitle)].filter(Boolean).join(" ");
+  let rank = tokenOverlap(cleanGoogleBooksTitle(originalTitle), candidateTitle) * 10;
+  if (expectedVolume) {
+    const candidateVolume = volumeNumber(candidateTitle);
+    rank += candidateVolume === expectedVolume ? 6 : candidateVolume ? -4 : 0;
+  }
+  if (originalAuthor && arrayJoin(info.authors)?.toLowerCase().includes(originalAuthor.toLowerCase())) rank += 2;
+  if (stringValue(info.description)) rank += 1;
+  if (coverUrl(info.imageLinks)) rank += 1;
+  return rank;
+}
+
+function volumeNumber(value: string): string | null {
+  return value.match(/\b(?:T|tome|vol(?:ume)?\.?)\s*0?(\d{1,3})\b/i)?.[1] ?? value.match(/\b0?(\d{1,3})\b/)?.[1] ?? null;
+}
+
+function tokenOverlap(left: string, right: string): number {
+  const leftTokens = new Set(normalizeTokens(left));
+  const rightTokens = new Set(normalizeTokens(right));
+  if (!leftTokens.size || !rightTokens.size) return 0;
+  return [...leftTokens].filter((token) => rightTokens.has(token)).length / leftTokens.size;
+}
+
+function normalizeTokens(value: string): string[] {
+  return value
+    .normalize("NFD")
+    .replace(/[\u0300-\u036f]/g, "")
+    .toLowerCase()
+    .replace(/[^a-z0-9]+/g, " ")
+    .split(" ")
+    .filter(Boolean);
+}
+
 function stringValue(value: unknown): string | null {
   return typeof value === "string" && value.trim() ? value.trim() : null;
 }
@@ -77,3 +182,20 @@ function isbnFromIndustryIdentifiers(value: unknown, expectedIsbn13: string | nu
   const isbn10 = entries.find((entry) => entry.type === "ISBN_10")?.identifier;
   return stringValue(isbn13) ?? stringValue(isbn10);
 }
+
+function isbnCandidates(value: unknown): string[] {
+  if (!Array.isArray(value)) return [];
+  return value.map((entry) => stringValue((entry as { identifier?: unknown }).identifier)).filter((entry): entry is string => Boolean(entry));
+}
+
+function coverUrl(value: unknown): string | null {
+  if (!value || typeof value !== "object") return null;
+  const links = value as Record<string, unknown>;
+  return (
+    stringValue(links.extraLarge) ??
+    stringValue(links.large) ??
+    stringValue(links.medium) ??
+    stringValue(links.thumbnail) ??
+    stringValue(links.smallThumbnail)
+  )?.replace(/^http:/, "https:") ?? null;
+}
diff --git a/apps/api/src/metadata/adapters/local.provider.ts b/apps/api/src/metadata/adapters/local.provider.ts
index 64061cf..8ac1762 100644
--- a/apps/api/src/metadata/adapters/local.provider.ts
+++ b/apps/api/src/metadata/adapters/local.provider.ts
@@ -1,5 +1,6 @@
 import { Injectable } from "@nestjs/common";
 import { MetadataLookup, MetadataMatch, MetadataProvider, MetadataProviderConfig, MetadataSearchQuery } from "../metadata.types.js";
+import { normalizePublishedDate } from "../use-cases/normalize-published-date.js";
 
 @Injectable()
 export class LocalMetadataProvider implements MetadataProvider {
@@ -20,7 +21,7 @@ export class LocalMetadataProvider implements MetadataProvider {
         title: query.title,
         author: query.author,
         isbn: query.isbn ?? null,
-        publishedDate: query.year ?? null
+        publishedDate: normalizePublishedDate(query.year)
       }
     ];
   }
diff --git a/apps/api/src/metadata/adapters/mangadex.provider.ts b/apps/api/src/metadata/adapters/mangadex.provider.ts
new file mode 100644
index 0000000..e8b3b2c
--- /dev/null
+++ b/apps/api/src/metadata/adapters/mangadex.provider.ts
@@ -0,0 +1,179 @@
+import { Injectable } from "@nestjs/common";
+import { MetadataLookup, MetadataMatch, MetadataProvider, MetadataProviderConfig, MetadataSearchQuery } from "../metadata.types.js";
+import { extractSeriesVolume } from "../use-cases/extract-series-volume.js";
+import { normalizePublishedDate } from "../use-cases/normalize-published-date.js";
+import { providerFetch } from "./provider-fetch.js";
+
+@Injectable()
+export class MangaDexProvider implements MetadataProvider {
+  readonly id = "mangadex" as const;
+
+  async lookup(lookup: MetadataLookup, config: MetadataProviderConfig): Promise<MetadataMatch | null> {
+    const matches = await this.searchByMetadata({ title: lookup.title, author: lookup.author, year: lookup.local.year }, config);
+    return matches[0] ?? null;
+  }
+
+  async searchByMetadata(query: MetadataSearchQuery, config: MetadataProviderConfig): Promise<MetadataMatch[]> {
+    const titles = mangaDexTitleQueries(query.title);
+    const matches: Array<MetadataMatch & { mangaDexRank?: number }> = [];
+    const seen = new Map<string, number>();
+    for (const title of titles) {
+      for (const match of await searchManga(title, config)) {
+        const key = match.sourceId ?? `${match.title}|${match.author}`;
+        const ranked = { ...match, scoreTitle: title, mangaDexRank: mangaRank(query.title, title, match) };
+        const existingIndex = seen.get(key);
+        if (existingIndex === undefined) {
+          seen.set(key, matches.length);
+          matches.push(ranked);
+          continue;
+        }
+        if ((ranked.mangaDexRank ?? 0) > (matches[existingIndex]?.mangaDexRank ?? 0)) {
+          matches[existingIndex] = ranked;
+        }
+      }
+    }
+    return rankMatches(matches);
+  }
+}
+
+export class MangaDexProviderError extends Error {
+  constructor(
+    readonly code: "rate-limit" | "http",
+    readonly status: number,
+    message: string
+  ) {
+    super(message);
+    this.name = "MangaDexProviderError";
+  }
+}
+
+async function searchManga(title: string, _config: MetadataProviderConfig): Promise<MetadataMatch[]> {
+  const url = new URL("https://api.mangadex.org/manga");
+  url.searchParams.set("title", title);
+  url.searchParams.set("limit", "10");
+  url.searchParams.set("includes[]", "cover_art");
+  url.searchParams.append("includes[]", "author");
+  url.searchParams.append("includes[]", "artist");
+  url.searchParams.set("contentRating[]", "safe");
+  url.searchParams.append("contentRating[]", "suggestive");
+  let response = await providerFetch("mangadex", url, {
+    headers: { "User-Agent": "ReadaBook/0.1 self-hosted metadata provider (MangaDex)" },
+    timeoutMs: 5000
+  });
+  if (response.status === 429) {
+    await sleep(retryDelayMs(response));
+    response = await providerFetch("mangadex", url, {
+      headers: { "User-Agent": "ReadaBook/0.1 self-hosted metadata provider (MangaDex)" },
+      timeoutMs: 5000
+    });
+  }
+  const data = (await parseMangaDexResponse(response)) as { data?: Array<Record<string, unknown>> };
+  return (data.data ?? []).map(mangaToMatch);
+}
+
+function retryDelayMs(response: Response): number {
+  const retryAfter = Number(response.headers.get("Retry-After"));
+  return Number.isFinite(retryAfter) && retryAfter > 0 ? Math.min(retryAfter * 1000, 2000) : 250;
+}
+
+function sleep(ms: number): Promise<void> {
+  return new Promise((resolve) => setTimeout(resolve, ms));
+}
+
+async function parseMangaDexResponse(response: Response): Promise<unknown> {
+  const data = (await response.json().catch(() => ({}))) as { errors?: Array<{ detail?: string; title?: string }> };
+  if (response.ok) return data;
+  const message = data.errors?.map((error) => error.detail ?? error.title).filter(Boolean).join("; ") || `MangaDex HTTP ${response.status}`;
+  if (response.status === 429) throw new MangaDexProviderError("rate-limit", response.status, message);
+  throw new MangaDexProviderError("http", response.status, message);
+}
+
+function mangaToMatch(manga: Record<string, unknown>): MetadataMatch & { mangaDexRank?: number } {
+  const id = stringValue(manga.id);
+  const attributes = objectValue(manga.attributes);
+  const relationships = Array.isArray(manga.relationships) ? (manga.relationships as Array<Record<string, unknown>>) : [];
+  const cover = relationships.find((entry) => entry.type === "cover_art");
+  const coverFile = stringValue(objectValue(cover?.attributes).fileName);
+  return {
+    title: localizedText(attributes.title) ?? undefined,
+    author: relationshipNames(relationships),
+    description: localizedText(attributes.description),
+    publishedDate: normalizePublishedDate(stringValue(attributes.year)),
+    language: stringValue(attributes.originalLanguage),
+    coverUrl: id && coverFile ? `https://uploads.mangadex.org/covers/${id}/${coverFile}.512.jpg` : null,
+    sourceId: id
+  };
+}
+
+function mangaDexTitleQueries(title: string): string[] {
+  const cleaned = extractSeriesVolume(title).seriesTitle;
+  return [cleaned, ...mangaDexTitleAliases(cleaned)].filter(
+    (value, index, values): value is string => Boolean(value) && values.indexOf(value) === index
+  );
+}
+
+function mangaDexTitleAliases(title: string): string[] {
+  const normalized = normalizeTitle(title);
+  if (normalized === "demon slayer school days") return ["Demon Slayer Kimetsu Academy", "Kimetsu Academy"];
+  return [];
+}
+
+function mangaRank(originalTitle: string, searchedTitle: string, match: MetadataMatch): number {
+  const expectedVolume = volumeNumber(originalTitle);
+  let rank = tokenOverlap(searchedTitle, match.title ?? "") * 10;
+  if (expectedVolume) {
+    const candidateVolume = volumeNumber(match.title ?? "");
+    rank += candidateVolume === expectedVolume ? 4 : candidateVolume ? -2 : 0;
+  }
+  if (match.coverUrl) rank += 1;
+  if (match.description) rank += 1;
+  return rank;
+}
+
+function rankMatches(matches: Array<MetadataMatch & { mangaDexRank?: number }>): MetadataMatch[] {
+  return [...matches].sort((left, right) => (right.mangaDexRank ?? 0) - (left.mangaDexRank ?? 0)).map(({ mangaDexRank: _rank, ...match }) => match);
+}
+
+function volumeNumber(value: string): string | null {
+  return value.match(/\b(?:T|tome|vol(?:ume)?\.?)\s*0?(\d{1,3})\b/i)?.[1] ?? value.match(/\b0?(\d{1,3})\b/)?.[1] ?? null;
+}
+
+function tokenOverlap(left: string, right: string): number {
+  const leftTokens = new Set(normalizeTitle(left).split(" ").filter(Boolean));
+  const rightTokens = new Set(normalizeTitle(right).split(" ").filter(Boolean));
+  if (!leftTokens.size || !rightTokens.size) return 0;
+  return [...leftTokens].filter((token) => rightTokens.has(token)).length / leftTokens.size;
+}
+
+function normalizeTitle(value: string): string {
+  return value
+    .normalize("NFD")
+    .replace(/[\u0300-\u036f]/g, "")
+    .toLowerCase()
+    .replace(/[^a-z0-9]+/g, " ")
+    .replace(/\s+/g, " ")
+    .trim();
+}
+
+function localizedText(value: unknown): string | null {
+  if (!value || typeof value !== "object") return null;
+  const entries = value as Record<string, unknown>;
+  return stringValue(entries.en) ?? stringValue(entries.fr) ?? Object.values(entries).map(stringValue).find(Boolean) ?? null;
+}
+
+function relationshipNames(relationships: Array<Record<string, unknown>>): string | null {
+  const names = relationships
+    .filter((entry) => entry.type === "author" || entry.type === "artist")
+    .map((entry) => stringValue(objectValue(entry.attributes).name))
+    .filter((entry): entry is string => Boolean(entry));
+  return names.length ? [...new Set(names)].join(", ") : null;
+}
+
+function objectValue(value: unknown): Record<string, unknown> {
+  return value && typeof value === "object" && !Array.isArray(value) ? (value as Record<string, unknown>) : {};
+}
+
+function stringValue(value: unknown): string | null {
+  if (typeof value === "number") return String(value);
+  return typeof value === "string" && value.trim() ? value.trim() : null;
+}
diff --git a/apps/api/src/metadata/adapters/open-library.provider.ts b/apps/api/src/metadata/adapters/open-library.provider.ts
index 892421f..2683b16 100644
--- a/apps/api/src/metadata/adapters/open-library.provider.ts
+++ b/apps/api/src/metadata/adapters/open-library.provider.ts
@@ -1,6 +1,8 @@
 import { Injectable } from "@nestjs/common";
 import { MetadataLookup, MetadataMatch, MetadataProvider, MetadataProviderConfig, MetadataSearchQuery } from "../metadata.types.js";
 import { toIsbn13 } from "../use-cases/extract-identifiers.js";
+import { normalizePublishedDate } from "../use-cases/normalize-published-date.js";
+import { providerFetch, providerHttpError } from "./provider-fetch.js";
 
 @Injectable()
 export class OpenLibraryProvider implements MetadataProvider {
@@ -24,11 +26,11 @@ export class OpenLibraryProvider implements MetadataProvider {
     if (query.author) url.searchParams.set("author", query.author);
     if (query.year) url.searchParams.set("first_publish_year", query.year);
     url.searchParams.set("limit", "5");
-    const response = await fetch(url, {
+    const response = await providerFetch(this.id, url, {
       headers: { "User-Agent": "ReadaBook/0.1 local-library-manager" },
-      signal: AbortSignal.timeout(4000)
+      timeoutMs: 4000
     });
-    if (!response.ok) return [];
+    if (!response.ok) throw await providerHttpError(this.id, response, `OpenLibrary HTTP ${response.status}`);
     const data = (await response.json()) as { docs?: Array<Record<string, unknown>> };
     return (data.docs ?? []).map((doc) => ({
       title: stringValue(doc.title) ?? undefined,
@@ -36,28 +38,31 @@ export class OpenLibraryProvider implements MetadataProvider {
       author: arrayJoin(doc.author_name),
       language: firstArrayValue(doc.language),
       publisher: firstArrayValue(doc.publisher),
-      publishedDate: String(doc.first_publish_year ?? "") || null,
-      isbn: bestIsbn(doc.isbn, query.isbn ? toIsbn13(query.isbn) : null)
+      publishedDate: normalizePublishedDate(String(doc.first_publish_year ?? "") || null),
+      isbn: bestIsbn(doc.isbn, query.isbn ? toIsbn13(query.isbn) : null),
+      coverUrl: openLibraryCoverUrl(doc.cover_i, firstArrayValue(doc.edition_key) ?? stringValue(doc.cover_edition_key))
     }));
   }
 
   private async lookupEdition(sourceId: string, expectedIsbn13: string | null): Promise<MetadataMatch | null> {
     const editionKey = sourceId.replace(/^\/?books\//, "");
     if (!editionKey) return null;
-    const response = await fetch(`https://openlibrary.org/books/${encodeURIComponent(editionKey)}.json`, {
+    const response = await providerFetch(this.id, `https://openlibrary.org/books/${encodeURIComponent(editionKey)}.json`, {
       headers: { "User-Agent": "ReadaBook/0.1 local-library-manager" },
-      signal: AbortSignal.timeout(4000)
+      timeoutMs: 4000
     });
-    if (!response.ok) return null;
+    if (response.status === 404) return null;
+    if (!response.ok) throw await providerHttpError(this.id, response, `OpenLibrary HTTP ${response.status}`);
     return this.editionToMatch((await response.json()) as Record<string, unknown>, expectedIsbn13);
   }
 
   private async lookupIsbn(isbn: string, expectedIsbn13: string | null): Promise<MetadataMatch | null> {
-    const response = await fetch(`https://openlibrary.org/isbn/${encodeURIComponent(isbn)}.json`, {
+    const response = await providerFetch(this.id, `https://openlibrary.org/isbn/${encodeURIComponent(isbn)}.json`, {
       headers: { "User-Agent": "ReadaBook/0.1 local-library-manager" },
-      signal: AbortSignal.timeout(4000)
+      timeoutMs: 4000
     });
-    if (!response.ok) return null;
+    if (response.status === 404) return null;
+    if (!response.ok) throw await providerHttpError(this.id, response, `OpenLibrary HTTP ${response.status}`);
     return this.editionToMatch((await response.json()) as Record<string, unknown>, expectedIsbn13);
   }
 
@@ -70,18 +75,20 @@ export class OpenLibraryProvider implements MetadataProvider {
       isbn: bestIsbn([...(asStringArray(edition.isbn_13)), ...(asStringArray(edition.isbn_10))], expectedIsbn13),
       language: languageValue(edition.languages),
       publisher: firstArrayValue(edition.publishers),
-      publishedDate: stringValue(edition.publish_date)
+      publishedDate: normalizePublishedDate(stringValue(edition.publish_date)),
+      coverUrl: editionCoverUrl(edition)
     };
   }
 
   private async lookupAuthorName(value: unknown): Promise<string | null> {
     const key = (Array.isArray(value) ? value[0] : undefined)?.key;
     if (typeof key !== "string") return null;
-    const response = await fetch(`https://openlibrary.org${key}.json`, {
+    const response = await providerFetch(this.id, `https://openlibrary.org${key}.json`, {
       headers: { "User-Agent": "ReadaBook/0.1 local-library-manager" },
-      signal: AbortSignal.timeout(3000)
+      timeoutMs: 3000
     });
-    if (!response.ok) return null;
+    if (response.status === 404) return null;
+    if (!response.ok) throw await providerHttpError(this.id, response, `OpenLibrary HTTP ${response.status}`);
     const author = (await response.json()) as Record<string, unknown>;
     return stringValue(author.name);
   }
@@ -124,3 +131,18 @@ function languageValue(value: unknown): string | null {
   const key = (Array.isArray(value) ? value[0] : undefined)?.key;
   return typeof key === "string" ? key.split("/").pop() ?? null : null;
 }
+
+function openLibraryCoverUrl(coverId: unknown, editionKey: string | null): string | null {
+  if (typeof coverId === "number" || typeof coverId === "string") {
+    return `https://covers.openlibrary.org/b/id/${encodeURIComponent(String(coverId))}-L.jpg`;
+  }
+  if (editionKey) {
+    return `https://covers.openlibrary.org/b/olid/${encodeURIComponent(editionKey)}-L.jpg`;
+  }
+  return null;
+}
+
+function editionCoverUrl(edition: Record<string, unknown>): string | null {
+  const covers = Array.isArray(edition.covers) ? edition.covers : [];
+  return openLibraryCoverUrl(covers[0], stringValue(edition.key)?.split("/").pop() ?? null);
+}
diff --git a/apps/api/src/metadata/adapters/provider-fetch.ts b/apps/api/src/metadata/adapters/provider-fetch.ts
new file mode 100644
index 0000000..f26c5fa
--- /dev/null
+++ b/apps/api/src/metadata/adapters/provider-fetch.ts
@@ -0,0 +1,88 @@
+import { MetadataProviderId } from "../metadata.types.js";
+
+export type MetadataProviderFailureCode = "timeout" | "dns" | "quota" | "auth" | "http" | "network";
+
+export class MetadataProviderRequestError extends Error {
+  constructor(
+    readonly provider: MetadataProviderId | "cover",
+    readonly code: MetadataProviderFailureCode,
+    readonly message: string,
+    readonly status?: number
+  ) {
+    super(message);
+    this.name = "MetadataProviderRequestError";
+  }
+}
+
+export async function providerFetch(
+  provider: MetadataProviderId | "cover",
+  input: string | URL,
+  init: RequestInit & { timeoutMs: number }
+): Promise<Response> {
+  const { timeoutMs, ...requestInit } = init;
+  try {
+    return await fetch(input, {
+      ...requestInit,
+      signal: requestInit.signal ?? AbortSignal.timeout(timeoutMs)
+    });
+  } catch (error) {
+    throw classifyFetchError(provider, error, timeoutMs);
+  }
+}
+
+export async function providerHttpError(
+  provider: MetadataProviderId | "cover",
+  response: Response,
+  fallbackMessage: string
+): Promise<MetadataProviderRequestError> {
+  const message = (await response.text().catch(() => "")) || fallbackMessage;
+  if (response.status === 429) return new MetadataProviderRequestError(provider, "quota", message, response.status);
+  if (response.status === 401 || response.status === 403) return new MetadataProviderRequestError(provider, "auth", message, response.status);
+  return new MetadataProviderRequestError(provider, "http", message, response.status);
+}
+
+export function describeMetadataProviderError(error: unknown): string {
+  if (error instanceof MetadataProviderRequestError) {
+    const status = error.status ? ` HTTP ${error.status}` : "";
+    return `${error.code}${status}: ${error.message}`;
+  }
+  if (hasProviderErrorCode(error)) {
+    const status = typeof error.status === "number" ? ` HTTP ${error.status}` : "";
+    return `${String(error.code)}${status}: ${errorMessage(error)}`;
+  }
+  return errorMessage(error);
+}
+
+function classifyFetchError(provider: MetadataProviderId | "cover", error: unknown, timeoutMs: number): MetadataProviderRequestError {
+  const code = nestedCode(error);
+  if (isTimeoutError(error)) {
+    return new MetadataProviderRequestError(provider, "timeout", `request timed out after ${timeoutMs}ms`);
+  }
+  if (code === "EAI_AGAIN" || code === "ENOTFOUND") {
+    return new MetadataProviderRequestError(provider, "dns", code);
+  }
+  return new MetadataProviderRequestError(provider, "network", errorMessage(error));
+}
+
+function isTimeoutError(error: unknown): boolean {
+  return (
+    error instanceof DOMException && (error.name === "AbortError" || error.name === "TimeoutError") ||
+    error instanceof Error && (error.name === "AbortError" || error.name === "TimeoutError")
+  );
+}
+
+function nestedCode(error: unknown): string | null {
+  if (!error || typeof error !== "object") return null;
+  const direct = "code" in error && typeof error.code === "string" ? error.code : null;
+  if (direct) return direct;
+  const cause = "cause" in error ? error.cause : null;
+  return cause && typeof cause === "object" && "code" in cause && typeof cause.code === "string" ? cause.code : null;
+}
+
+function hasProviderErrorCode(error: unknown): error is { code: string; status?: number; message?: string } {
+  return Boolean(error && typeof error === "object" && "code" in error && typeof error.code === "string");
+}
+
+function errorMessage(error: unknown): string {
+  return error instanceof Error ? error.message : String(error);
+}
diff --git a/apps/api/src/metadata/local-metadata-hints.test.ts b/apps/api/src/metadata/local-metadata-hints.test.ts
index 19f9e2b..53bab86 100644
--- a/apps/api/src/metadata/local-metadata-hints.test.ts
+++ b/apps/api/src/metadata/local-metadata-hints.test.ts
@@ -38,6 +38,76 @@ describe("metadata match scoring", () => {
     );
 
     expect(best?.match.title).toBe("Harry Potter et le Prince de sang-mêlé");
-    expect(best?.score).toBeGreaterThan(0.7);
+    expect(best?.score).toBeGreaterThan(70);
+  });
+
+  it("scores titles, authors and dates with the contract weights", () => {
+    const scorer = new ScoreMetadataMatch();
+    const result = scorer.details(
+      { title: "The Harry Potter et le prince de sang mêlé: édition collector", author: "J. K. Rowling", year: "2005" },
+      {
+        title: "Harry Potter et le prince de sang-mêlé",
+        author: "J.K. Rowling",
+        publishedDate: "2006"
+      }
+    );
+
+    expect(result.titleScore).toBe(100);
+    expect(result.authorScore).toBe(30);
+    expect(result.dateScore).toBe(5);
+    expect(result.score).toBe(96);
+  });
+
+  it("uses exact ISBN matches before weaker title-only candidates", () => {
+    const best = new ScoreMetadataMatch().best(
+      { title: "Daredevil", author: null, isbn: "9782809476255" },
+      [
+        {
+          title: "Daredevil",
+          author: "Rosemary Carter",
+          isbn: "9780373105601"
+        },
+        {
+          title: "Daredevil by Chip Zdarsky",
+          author: "Chip Zdarsky",
+          isbn: "9782809476255"
+        }
+      ]
+    );
+
+    expect(best?.match.author).toBe("Chip Zdarsky");
+    expect(best?.isbnMatch).toBe(true);
+  });
+
+  it("scores unrelated serialized or audiobook candidates from title/author/date only", () => {
+    const scorer = new ScoreMetadataMatch();
+    const query = { title: "Harry Potter et le prince de sang mêlé", author: "J. K. Rowling" };
+    const french = scorer.score(query, {
+      title: "Harry Potter et le prince de sang-mêlé",
+      author: "J. K. Rowling",
+      isbn: "9782070577644"
+    });
+    const koreanVolume = scorer.score(query, {
+      title: "Harry Potter et le prince de sang-mêlé - Volume 1",
+      author: "J. K. Rowling",
+      publisher: "문학수첩",
+      isbn: "9791193790724"
+    });
+
+    expect(french).toBeGreaterThan(80);
+    expect(koreanVolume).toBeLessThan(french);
+  });
+
+  it("does not apply legacy audiobook penalties outside the contract", () => {
+    const scorer = new ScoreMetadataMatch();
+    const query = { title: "Harry Potter et le prince de sang mêlé", author: "J. K. Rowling" };
+
+    expect(
+      scorer.score(query, {
+        title: "Harry Potter Et Le Prince De Sang-mêlé Livre Audio",
+        author: "J. K. Rowling",
+        isbn: "9782075105170"
+      })
+    ).toBeGreaterThan(80);
   });
 });
diff --git a/apps/api/src/metadata/metadata-providers.test.ts b/apps/api/src/metadata/metadata-providers.test.ts
index da905c1..df6cc28 100644
--- a/apps/api/src/metadata/metadata-providers.test.ts
+++ b/apps/api/src/metadata/metadata-providers.test.ts
@@ -1,7 +1,10 @@
 import { afterEach, describe, expect, it, vi } from "vitest";
 import { BnfProvider } from "./adapters/bnf.provider.js";
-import { GoogleBooksProvider } from "./adapters/google-books.provider.js";
+import { ComicVineProvider } from "./adapters/comic-vine.provider.js";
+import { GoogleBooksProvider, GoogleBooksProviderError } from "./adapters/google-books.provider.js";
+import { MangaDexProvider } from "./adapters/mangadex.provider.js";
 import { OpenLibraryProvider } from "./adapters/open-library.provider.js";
+import { MetadataProviderRequestError } from "./adapters/provider-fetch.js";
 import { MetadataLookup, MetadataProviderConfig } from "./metadata.types.js";
 
 const lookup: MetadataLookup = {
@@ -58,14 +61,210 @@ describe("metadata providers", () => {
     });
   });
 
-  it("treats Google Books quota exhaustion as a non-blocking miss", async () => {
+  it("reports Google Books quota exhaustion explicitly", async () => {
     const fetchMock = vi.fn(async () => jsonResponse({ error: { code: 429, status: "RESOURCE_EXHAUSTED" } }, 429));
     vi.stubGlobal("fetch", fetchMock);
 
-    const result = await new GoogleBooksProvider().lookup(lookup, { ...config, provider: "googlebooks" });
+    await expect(new GoogleBooksProvider().lookup(lookup, { ...config, provider: "googlebooks" })).rejects.toMatchObject({
+      code: "quota",
+      status: 429
+    } satisfies Partial<GoogleBooksProviderError>);
 
     expect(String((fetchMock.mock.calls[0] as unknown[])[0])).toContain("q=isbn%3A9782070612376");
-    expect(result).toBeNull();
+  });
+
+  it("classifies provider DNS failures explicitly", async () => {
+    const error = new TypeError("fetch failed") as Error & { cause?: { code: string } };
+    error.cause = { code: "EAI_AGAIN" };
+    vi.stubGlobal("fetch", vi.fn(async () => Promise.reject(error)));
+
+    await expect(new OpenLibraryProvider().searchByMetadata({ title: "Daredevil", author: null }, config)).rejects.toMatchObject({
+      code: "dns",
+      message: "EAI_AGAIN"
+    } satisfies Partial<MetadataProviderRequestError>);
+  });
+
+  it.each([
+    ["Demon.Slayer.School.Days.T01.FRENCH.CBZ.eBook-ebdz", "Demon Slayer School Days 1"],
+    ["Dragon.Ball.SD.T01.FRENCH.CBZ.eBook-Paprika+", "Dragon Ball SD 1"],
+    ["Eyeshield.21.T01.FRENCH.CBZ.eBook-ebdz", "Eyeshield 21 1"],
+    ["Solo Leveling T03", "Solo Leveling 3"]
+  ])("cleans noisy Google Books title queries for %s", async (title, expectedCleanTitle) => {
+    const fetchMock = vi.fn(async () => jsonResponse({ totalItems: 0, items: [] }));
+    vi.stubGlobal("fetch", fetchMock);
+
+    await new GoogleBooksProvider().searchByMetadata({ title, author: null }, { ...config, provider: "googlebooks" });
+    const queries = fetchMock.mock.calls.map((call) => new URL(String((call as unknown[])[0])).searchParams.get("q") ?? "");
+
+    expect(queries[0]).toBe(`intitle:"${expectedCleanTitle}"`);
+    expect(queries.join(" ")).not.toMatch(/\b(FRENCH|CBZ|eBook|ebdz|Paprika)\b/i);
+  });
+
+  it("sorts Google Books results by matching manga volume instead of taking the first item", async () => {
+    const fetchMock = vi.fn(async () =>
+      jsonResponse({
+        totalItems: 2,
+        items: [
+          { id: "volume-2", volumeInfo: { title: "Solo Leveling, Vol. 2", authors: ["Chugong"], publishedDate: "2021" } },
+          { id: "volume-3", volumeInfo: { title: "Solo Leveling, Vol. 3", authors: ["Chugong"], publishedDate: "2021" } }
+        ]
+      })
+    );
+    vi.stubGlobal("fetch", fetchMock);
+
+    const results = await new GoogleBooksProvider().searchByMetadata(
+      { title: "Solo Leveling T03", author: null },
+      { ...config, provider: "googlebooks" }
+    );
+
+    expect(results[0]).toMatchObject({ title: "Solo Leveling, Vol. 3", sourceId: "volume-3" });
+  });
+
+  it.each([
+    ["Solo Leveling T03", "Solo Leveling"],
+    ["Solo Leveling 003", "Solo Leveling"],
+    ["Eyeshield.21.T01.FRENCH.CBZ.eBook-ebdz", "Eyeshield 21"],
+    ["Dragon.Ball.SD.T01.FRENCH.CBZ.eBook-Paprika+", "Dragon Ball SD"]
+  ])("queries MangaDex with the cleaned series title for %s", async (title, expectedQuery) => {
+    const fetchMock = vi.fn(async () => jsonResponse({ data: [] }));
+    vi.stubGlobal("fetch", fetchMock);
+
+    await new MangaDexProvider().searchByMetadata({ title, author: null }, { ...config, provider: "mangadex" });
+    const firstUrl = new URL(String((fetchMock.mock.calls[0] as unknown[])[0]));
+
+    expect(firstUrl.searchParams.get("title")).toBe(expectedQuery);
+  });
+
+  it("queries MangaDex aliases and maps cover_art to a cover URL", async () => {
+    const fetchMock = vi
+      .fn()
+      .mockResolvedValueOnce(jsonResponse({ data: [] }))
+      .mockResolvedValueOnce(
+        jsonResponse({
+          data: [
+            {
+              id: "manga-1",
+              attributes: {
+                title: { en: "Demon Slayer: Kimetsu Academy" },
+                description: { en: "School spin-off." },
+                year: 2021,
+                originalLanguage: "ja"
+              },
+              relationships: [
+                { type: "cover_art", attributes: { fileName: "cover.jpg" } },
+                { type: "author", attributes: { name: "Natsuki Hokami" } }
+              ]
+            }
+          ]
+        })
+      )
+      .mockResolvedValue(jsonResponse({ data: [] }));
+    vi.stubGlobal("fetch", fetchMock);
+
+    const results = await new MangaDexProvider().searchByMetadata(
+      { title: "Demon.Slayer.School.Days.T01.FRENCH.CBZ.eBook-ebdz", author: null },
+      { ...config, provider: "mangadex" }
+    );
+    const firstUrl = new URL(String((fetchMock.mock.calls[0] as unknown[])[0]));
+    const secondUrl = new URL(String((fetchMock.mock.calls[1] as unknown[])[0]));
+
+    expect(firstUrl.searchParams.get("title")).toBe("Demon Slayer School Days");
+    expect(secondUrl.searchParams.get("title")).toBe("Demon Slayer Kimetsu Academy");
+    expect(results[0]).toMatchObject({
+      title: "Demon Slayer: Kimetsu Academy",
+      scoreTitle: "Demon Slayer Kimetsu Academy",
+      author: "Natsuki Hokami",
+      publishedDate: "2021",
+      coverUrl: "https://uploads.mangadex.org/covers/manga-1/cover.jpg.512.jpg"
+    });
+  });
+
+  it("keeps Dragon Ball SD ahead of Dragon Ball for MangaDex matches", async () => {
+    const fetchMock = vi.fn(async () =>
+      jsonResponse({
+        data: [
+          {
+            id: "dragon-ball",
+            attributes: { title: { en: "Dragon Ball" }, description: { en: "Original series." }, year: 1984, originalLanguage: "ja" },
+            relationships: []
+          },
+          {
+            id: "dragon-ball-sd",
+            attributes: { title: { en: "Dragon Ball SD" }, description: { en: "SD spin-off." }, year: 2010, originalLanguage: "ja" },
+            relationships: []
+          }
+        ]
+      })
+    );
+    vi.stubGlobal("fetch", fetchMock);
+
+    const results = await new MangaDexProvider().searchByMetadata(
+      { title: "Dragon.Ball.SD.T01.FRENCH.CBZ.eBook-Paprika+", author: null },
+      { ...config, provider: "mangadex" }
+    );
+
+    expect(results[0]).toMatchObject({ title: "Dragon Ball SD", sourceId: "dragon-ball-sd" });
+  });
+
+  it("reports MangaDex rate limits explicitly", async () => {
+    vi.stubGlobal("fetch", vi.fn(async () => jsonResponse({ errors: [{ detail: "Too many requests" }] }, 429)));
+
+    await expect(
+      new MangaDexProvider().searchByMetadata({ title: "Solo Leveling T03", author: null }, { ...config, provider: "mangadex" })
+    ).rejects.toMatchObject({ code: "rate-limit", status: 429 });
+  });
+
+  it("requires a Comic Vine API key before querying", async () => {
+    await expect(
+      new ComicVineProvider().searchByMetadata({ title: "Wolverine Origin", author: null }, { ...config, provider: "comicvine", apiKey: null })
+    ).rejects.toMatchObject({ code: "missing-key" });
+  });
+
+  it("queries Comic Vine volumes/issues and cleans HTML descriptions", async () => {
+    const fetchMock = vi.fn(async () =>
+      jsonResponse({
+        status_code: 1,
+        results: [
+          {
+            id: 123,
+            name: "Wolverine: The Origin",
+            description: "<p>Origin story &amp; family secrets.</p>",
+            start_year: "2001",
+            image: { super_url: "https://comicvine.gamespot.com/a/uploads/scale_large/origin.jpg" },
+            publisher: { name: "Marvel" }
+          }
+        ]
+      })
+    );
+    vi.stubGlobal("fetch", fetchMock);
+
+    const results = await new ComicVineProvider().searchByMetadata(
+      { title: "Comics.Fr.Wolverine.Origin.by.AleK.(emuleCenter.net)", author: null },
+      { ...config, provider: "comicvine", apiKey: "cv-key" }
+    );
+    const firstUrl = new URL(String((fetchMock.mock.calls[0] as unknown[])[0]));
+
+    expect(firstUrl.searchParams.get("resources")).toBe("volume");
+    expect(firstUrl.searchParams.get("query")).toBe("Comics Fr Wolverine Origin by AleK");
+    expect(results[0]).toMatchObject({
+      title: "Wolverine: The Origin",
+      description: "Origin story & family secrets.",
+      publishedDate: "2001",
+      publisher: "Marvel",
+      coverUrl: "https://comicvine.gamespot.com/a/uploads/scale_large/origin.jpg"
+    });
+  });
+
+  it("reports Comic Vine invalid keys and rate limits explicitly", async () => {
+    vi.stubGlobal("fetch", vi.fn(async () => jsonResponse({ status_code: 101, error: "Invalid API Key" })));
+    await expect(
+      new ComicVineProvider().searchByMetadata({ title: "Daredevil", author: null }, { ...config, provider: "comicvine", apiKey: "bad" })
+    ).rejects.toMatchObject({ code: "invalid-key" });
+
+    vi.stubGlobal("fetch", vi.fn(async () => jsonResponse({ error: "Rate limited" }, 429)));
+    await expect(
+      new ComicVineProvider().searchByMetadata({ title: "Daredevil", author: null }, { ...config, provider: "comicvine", apiKey: "ok" })
+    ).rejects.toMatchObject({ code: "rate-limit", status: 429 });
   });
 
   it("queries OpenLibrary by local metadata when ISBN is missing", async () => {
diff --git a/apps/api/src/metadata/metadata.module.ts b/apps/api/src/metadata/metadata.module.ts
index d15002a..5f7145b 100644
--- a/apps/api/src/metadata/metadata.module.ts
+++ b/apps/api/src/metadata/metadata.module.ts
@@ -1,14 +1,16 @@
 import { Module } from "@nestjs/common";
 import { DatabaseModule } from "../database/database.module.js";
 import { BnfProvider } from "./adapters/bnf.provider.js";
+import { ComicVineProvider } from "./adapters/comic-vine.provider.js";
 import { GoogleBooksProvider } from "./adapters/google-books.provider.js";
 import { LocalMetadataProvider } from "./adapters/local.provider.js";
+import { MangaDexProvider } from "./adapters/mangadex.provider.js";
 import { OpenLibraryProvider } from "./adapters/open-library.provider.js";
 import { MetadataService } from "./metadata.service.js";
 
 @Module({
   imports: [DatabaseModule],
-  providers: [MetadataService, LocalMetadataProvider, OpenLibraryProvider, GoogleBooksProvider, BnfProvider],
+  providers: [MetadataService, LocalMetadataProvider, OpenLibraryProvider, GoogleBooksProvider, BnfProvider, MangaDexProvider, ComicVineProvider],
   exports: [MetadataService]
 })
 export class MetadataModule {}
diff --git a/apps/api/src/metadata/metadata.service.test.ts b/apps/api/src/metadata/metadata.service.test.ts
index 0e1adea..9a7823d 100644
--- a/apps/api/src/metadata/metadata.service.test.ts
+++ b/apps/api/src/metadata/metadata.service.test.ts
@@ -1,9 +1,12 @@
-import { mkdtempSync, rmSync } from "node:fs";
+import { existsSync, mkdirSync, mkdtempSync, rmSync } from "node:fs";
 import { tmpdir } from "node:os";
 import { join } from "node:path";
+import AdmZip from "adm-zip";
 import { afterEach, describe, expect, it, vi } from "vitest";
 import { DatabaseService } from "../database/database.service.js";
+import { books, libraries } from "../database/schema.js";
 import { BookMetadata } from "../scanner/metadata.js";
+import { MetadataProviderRequestError } from "./adapters/provider-fetch.js";
 import { MetadataService } from "./metadata.service.js";
 import { MetadataLookup, MetadataMatch, MetadataProvider, MetadataProviderConfig, MetadataSearchQuery } from "./metadata.types.js";
 
@@ -52,7 +55,9 @@ describe("MetadataService", () => {
       localProvider as never,
       openLibraryProvider as never,
       providerStub("googlebooks") as never,
-      providerStub("bnf") as never
+      providerStub("bnf") as never,
+      providerStub("mangadex") as never,
+      providerStub("comicvine") as never
     );
     const localMetadata: BookMetadata = {
       title: "Harry Potter et le Prince de Sang Mele",
@@ -73,7 +78,7 @@ describe("MetadataService", () => {
     expect(openLibraryLookup).toHaveBeenCalledOnce();
     expect(openLibraryLookup.mock.calls[0]?.[0].identifiers.isbn13).toBe("9782070612383");
     expect(result).toMatchObject({
-      title: "Harry Potter et le Prince de sang-mêlé",
+      title: "Harry Potter et le Prince de Sang Mele",
       isbn: "9782070612383",
       isbn13: "9782070612383",
       description: "Harry Potter découvre l'héritage du Prince de Sang-Mêlé."
@@ -82,6 +87,64 @@ describe("MetadataService", () => {
     database.onModuleDestroy();
   });
 
+  it.runIf(canLoadBetterSqlite())("continues enrichment after a provider DNS failure and logs the failure class", async () => {
+    const database = createDatabase();
+    database.sqlite.prepare("UPDATE metadata_source_config SET enabled = 1 WHERE provider = 'googlebooks'").run();
+    const warn = vi.spyOn(console, "warn").mockImplementation(() => undefined);
+    const openLibraryProvider: MetadataProvider = {
+      id: "openlibrary",
+      lookup: async () => null,
+      searchByMetadata: async () => {
+        throw new MetadataProviderRequestError("openlibrary", "dns", "EAI_AGAIN");
+      }
+    };
+    const googleProvider: MetadataProvider = {
+      id: "googlebooks",
+      lookup: async () => null,
+      searchByMetadata: async () => [
+        {
+          title: "Daredevil",
+          author: "Roy Thomas",
+          description: "Daredevil keeps moving even when another provider is unreachable.",
+          publishedDate: "2019"
+        }
+      ]
+    };
+    const service = new MetadataService(
+      database,
+      providerStub("local") as never,
+      openLibraryProvider as never,
+      googleProvider as never,
+      providerStub("bnf") as never,
+      providerStub("mangadex") as never,
+      providerStub("comicvine") as never
+    );
+
+    const result = await service.enrichMetadata(
+      {
+        title: "Daredevil",
+        author: null,
+        description: null,
+        isbn: null,
+        language: null,
+        publisher: null,
+        publishedDate: null,
+        coverPath: null
+      },
+      "/library/Daredevil.cbz",
+      { remote: true }
+    );
+
+    expect(warn).toHaveBeenCalledWith('[metadata] Provider openlibrary failed for "Daredevil": dns: EAI_AGAIN');
+    expect(result).toMatchObject({
+      title: "Daredevil",
+      author: "Roy Thomas",
+      description: "Daredevil keeps moving even when another provider is unreachable."
+    });
+
+    database.onModuleDestroy();
+  });
+
   it.runIf(canLoadBetterSqlite())("looks up details from a title search hit even when the hit has no ISBN", async () => {
     const database = createDatabase();
     const localProvider = providerStub("local");
@@ -112,7 +175,9 @@ describe("MetadataService", () => {
       localProvider as never,
       openLibraryProvider as never,
       providerStub("googlebooks") as never,
-      providerStub("bnf") as never
+      providerStub("bnf") as never,
+      providerStub("mangadex") as never,
+      providerStub("comicvine") as never
     );
     const localMetadata: BookMetadata = {
       title: "Harry Potter et le prince de sang mele",
@@ -133,7 +198,7 @@ describe("MetadataService", () => {
     expect(openLibraryLookup).toHaveBeenCalledOnce();
     expect(openLibraryLookup.mock.calls[0]?.[0].sourceId).toBe("OL24333986M");
     expect(result).toMatchObject({
-      title: "Harry Potter et le prince de sang-mêlé",
+      title: "Harry Potter et le prince de sang mele",
       author: "J. K. Rowling",
       isbn: "9782070612383",
       isbn13: "9782070612383",
@@ -143,7 +208,61 @@ describe("MetadataService", () => {
     database.onModuleDestroy();
   });
 
-  it.runIf(canLoadBetterSqlite())("replaces an ambiguous title-only identification when a later provider supplies a described record", async () => {
+  it.runIf(canLoadBetterSqlite())("backfills a missing ISBN lookup description from a high-confidence title search", async () => {
+    const database = createDatabase();
+    const openLibraryProvider: MetadataProvider = {
+      id: "openlibrary",
+      lookup: async () => ({
+        title: "Harry Potter et la coupe de feu",
+        author: "J. K. Rowling",
+        isbn: "9782070624553",
+        publisher: "Gallimard",
+        publishedDate: "2016"
+      }),
+      searchByMetadata: async () => [
+        {
+          title: "Harry Potter et la coupe de feu",
+          author: "J. K. Rowling",
+          isbn: "9782070619207",
+          description: "Harry est invité à assister à la Coupe du monde de Quidditch."
+        }
+      ]
+    };
+    const service = new MetadataService(
+      database,
+      providerStub("local") as never,
+      openLibraryProvider as never,
+      providerStub("googlebooks") as never,
+      providerStub("bnf") as never,
+      providerStub("mangadex") as never,
+      providerStub("comicvine") as never
+    );
+
+    const result = await service.enrichMetadata(
+      {
+        title: "Harry Potter et la coupe de feu",
+        author: "J. K. Rowling",
+        description: null,
+        isbn: "9782070624553",
+        language: null,
+        publisher: null,
+        publishedDate: null,
+        coverPath: "/covers/local.jpg"
+      },
+      "/library/Harry Potter et la coupe de feu.epub",
+      { remote: true }
+    );
+
+    expect(result).toMatchObject({
+      isbn: "9782070624553",
+      description: "Harry est invité à assister à la Coupe du monde de Quidditch.",
+      coverPath: "/covers/local.jpg"
+    });
+
+    database.onModuleDestroy();
+  });
+
+  it.runIf(canLoadBetterSqlite())("uses provider priority as the tie-breaker for high-confidence title matches", async () => {
     const database = createDatabase();
     database.sqlite.prepare("UPDATE metadata_source_config SET enabled = 1 WHERE provider = 'bnf'").run();
     const openLibraryProvider: MetadataProvider = {
@@ -179,7 +298,9 @@ describe("MetadataService", () => {
       providerStub("local") as never,
       openLibraryProvider as never,
       providerStub("googlebooks") as never,
-      bnfProvider as never
+      bnfProvider as never,
+      providerStub("mangadex") as never,
+      providerStub("comicvine") as never
     );
     const localMetadata: BookMetadata = {
       title: "Daredevil",
@@ -196,14 +317,583 @@ describe("MetadataService", () => {
 
     expect(result).toMatchObject({
       title: "Daredevil",
-      author: "scénario, Roy Thomas, Gary Friedrich",
-      isbn: "9782809476255",
-      isbn13: "9782809476255",
+      author: "Rosemary Carter",
+      isbn: "9780373105601",
+      isbn13: "9780373105601",
       description: "Daredevil affronte l'Homme aux échasses."
     });
 
     database.onModuleDestroy();
   });
+
+  it.runIf(canLoadBetterSqlite())("re-enriches from stored local hints instead of a previously failed remote ISBN", async () => {
+    const database = createDatabase();
+    const now = database.now();
+    const library = database.db
+      .insert(libraries)
+      .values({ name: "Comics", path: "/library", enabled: true, createdAt: now, updatedAt: now })
+      .returning()
+      .get();
+    const localMetadataJson = JSON.stringify({
+      title: "Daredevil",
+      author: null,
+      year: null,
+      isbn: null,
+      fileTitle: "Daredevil",
+      raw: { title: "Daredevil", author: null, publishedDate: null, fileName: "Daredevil" }
+    });
+    const book = database.db
+      .insert(books)
+      .values({
+        libraryId: library.id,
+        title: "Daredevil",
+        author: "Rosemary Carter",
+        description: null,
+        isbn: "9780373105601",
+        isbn13: "9780373105601",
+        identifiersJson: JSON.stringify({ isbn10: null, isbn13: null, candidates: [] }),
+        localMetadataJson,
+        language: null,
+        publisher: "Harlequin Books",
+        publishedDate: "1982",
+        format: "cbz",
+        filePath: "/library/Daredevil.cbz",
+        coverPath: "/covers/daredevil.jpg",
+        scanStatus: "succeeded",
+        enrichmentStatus: "failed",
+        fileSize: 42,
+        fileMtime: now,
+        createdAt: now,
+        updatedAt: now
+      })
+      .returning()
+      .get();
+    const openLibraryLookup = vi.fn<(_: MetadataLookup, __: MetadataProviderConfig) => Promise<MetadataMatch | null>>(async () => null);
+    const bnfSearch = vi.fn<(_: MetadataSearchQuery, __: MetadataProviderConfig) => Promise<MetadataMatch[]>>(async () => [
+      {
+        title: "Daredevil",
+        author: "scénario, Roy Thomas, Gary Friedrich",
+        isbn: "9782809476255",
+        description: "Daredevil affronte l'Homme aux échasses.",
+        publisher: "Panini comics",
+        publishedDate: "2019"
+      }
+    ]);
+    database.sqlite.prepare("UPDATE metadata_source_config SET enabled = 1 WHERE provider = 'bnf'").run();
+    const service = new MetadataService(
+      database,
+      providerStub("local") as never,
+      { id: "openlibrary", lookup: openLibraryLookup, searchByMetadata: async () => [] } as never,
+      providerStub("googlebooks") as never,
+      { id: "bnf", lookup: async () => null, searchByMetadata: bnfSearch } as never,
+      providerStub("mangadex") as never,
+      providerStub("comicvine") as never
+    );
+
+    const result = await service.enrichBook(book.id);
+
+    expect(openLibraryLookup).not.toHaveBeenCalled();
+    expect(bnfSearch.mock.calls[0]?.[0].title).toBe("Daredevil");
+    expect(result).toMatchObject({
+      author: "Rosemary Carter",
+      isbn: "9780373105601",
+      isbn13: "9780373105601",
+      publisher: "Harlequin Books",
+      coverPath: "/covers/daredevil.jpg"
+    });
+
+    database.onModuleDestroy();
+  });
+
+  it.runIf(canLoadBetterSqlite())("stores provider covers as local bytes and exposes field provenance with metadata status", async () => {
+    const database = createDatabase();
+    const fetchMock = vi.spyOn(globalThis, "fetch").mockResolvedValue({
+      ok: true,
+      headers: new Headers({ "content-type": "image/jpeg" }),
+      arrayBuffer: async () => new Uint8Array([1, 2, 3, 4]).buffer
+    } as Response);
+    const openLibraryProvider: MetadataProvider = {
+      id: "openlibrary",
+      lookup: async () => null,
+      searchByMetadata: async () => [
+        {
+          title: "Daredevil",
+          author: "Roy Thomas",
+          description: "Daredevil affronte une nouvelle menace.",
+          isbn: "9782809476255",
+          coverUrl: "https://covers.openlibrary.org/b/id/123-L.jpg"
+        }
+      ]
+    };
+    const service = new MetadataService(
+      database,
+      providerStub("local") as never,
+      openLibraryProvider as never,
+      providerStub("googlebooks") as never,
+      providerStub("bnf") as never,
+      providerStub("mangadex") as never,
+      providerStub("comicvine") as never
+    );
+
+    const result = await service.enrichMetadata(
+      {
+        title: "Daredevil",
+        author: null,
+        description: null,
+        isbn: null,
+        language: null,
+        publisher: null,
+        publishedDate: null,
+        coverPath: null
+      },
+      "/library/Daredevil.cbz",
+      { remote: true }
+    );
+
+    expect(fetchMock).toHaveBeenCalledWith("https://covers.openlibrary.org/b/id/123-L.jpg", expect.any(Object));
+    expect(result.coverPath).toMatch(/storage\/covers\/.+\.jpg$/);
+    expect(result.metadataStatus).toBe("enriched");
+    expect(JSON.parse(result.metadataProvenanceJson)).toMatchObject({
+      title: "local",
+      author: "openlibrary",
+      description: "openlibrary",
+      coverPath: "openlibrary"
+    });
+
+    database.onModuleDestroy();
+  });
+
+  it.runIf(canLoadBetterSqlite())("retrofits a local cover for an existing book during metadata enrichment", async () => {
+    const database = createDatabase();
+    mkdirSync(database.config.storageDir, { recursive: true });
+    const filePath = join(database.config.storageDir, "Demon.Slayer.School.Days.T01.FRENCH.CBZ");
+    const zip = new AdmZip();
+    zip.addFile("001.jpg", Buffer.from([0xff, 0xd8, 0xff, 0xd9]));
+    zip.writeZip(filePath);
+    const now = database.now();
+    const library = database.db
+      .insert(libraries)
+      .values({ name: "Comics", path: database.config.storageDir, enabled: true, createdAt: now, updatedAt: now })
+      .returning()
+      .get();
+    const book = database.db
+      .insert(books)
+      .values({
+        libraryId: library.id,
+        title: "Demon.Slayer.School.Days.T01.FRENCH.CBZ. -ebdz",
+        author: null,
+        description: null,
+        isbn: null,
+        isbn13: null,
+        identifiersJson: JSON.stringify({ isbn10: null, isbn13: null, candidates: [] }),
+        localMetadataJson: JSON.stringify({
+          title: "Demon Slayer School Days T01 FRENCH",
+          author: null,
+          year: null,
+          isbn: null,
+          fileTitle: "Demon Slayer School Days T01 FRENCH",
+          raw: { title: "Demon.Slayer.School.Days.T01.FRENCH.CBZ. -ebdz", author: null, publishedDate: null, fileName: "Demon.Slayer.School.Days.T01.FRENCH.CBZ. -ebdz" }
+        }),
+        language: null,
+        publisher: null,
+        publishedDate: null,
+        format: "cbz",
+        filePath,
+        coverPath: null,
+        metadataStatus: "none",
+        metadataProvenanceJson: JSON.stringify({ title: "local" }),
+        scanStatus: "succeeded",
+        enrichmentStatus: "succeeded",
+        fileSize: 42,
+        fileMtime: now,
+        createdAt: now,
+        updatedAt: now
+      })
+      .returning()
+      .get();
+    const service = new MetadataService(
+      database,
+      providerStub("local") as never,
+      providerStub("openlibrary") as never,
+      providerStub("googlebooks") as never,
+      providerStub("bnf") as never,
+      providerStub("mangadex") as never,
+      providerStub("comicvine") as never
+    );
+
+    const result = await service.enrichBook(book.id);
+
+    expect(result.coverPath).toMatch(/covers\/[a-f0-9]+\.jpg$/);
+    expect(result.coverPath && existsSync(result.coverPath)).toBe(true);
+    expect(result.metadataStatus).toBe("partial");
+    expect(JSON.parse(result.metadataProvenanceJson ?? "{}")).toMatchObject({ coverPath: "local" });
+
+    database.onModuleDestroy();
+  });
+
+  it.runIf(canLoadBetterSqlite())("drops sentinel publication dates from provider matches for real affected titles", async () => {
+    const database = createDatabase();
+    const openLibraryProvider: MetadataProvider = {
+      id: "openlibrary",
+      lookup: async () => null,
+      searchByMetadata: async () => [
+        {
+          title: "Harry Potter et les reliques de la mort",
+          author: "J. K. Rowling",
+          publishedDate: "0101-01-01T00:00:00+00:00",
+          description: "Septième année."
+        }
+      ]
+    };
+    const service = new MetadataService(
+      database,
+      providerStub("local") as never,
+      openLibraryProvider as never,
+      providerStub("googlebooks") as never,
+      providerStub("bnf") as never,
+      providerStub("mangadex") as never,
+      providerStub("comicvine") as never
+    );
+
+    const result = await service.enrichMetadata(
+      {
+        title: "Harry Potter et les reliques de la mort",
+        author: "J. K. Rowling",
+        description: null,
+        isbn: null,
+        language: null,
+        publisher: null,
+        publishedDate: null,
+        coverPath: null
+      },
+      "/library/Harry Potter et les reliques de la mort.epub",
+      { remote: true }
+    );
+
+    expect(result.publishedDate).toBeNull();
+
+    database.onModuleDestroy();
+  });
+
+  it.runIf(canLoadBetterSqlite())("does not overwrite an existing valid date with a provider sentinel", async () => {
+    const database = createDatabase();
+    const now = database.now();
+    const library = database.db
+      .insert(libraries)
+      .values({ name: "Novels", path: "/library", enabled: true, createdAt: now, updatedAt: now })
+      .returning()
+      .get();
+    const book = database.db
+      .insert(books)
+      .values({
+        libraryId: library.id,
+        title: "Lord of the Mysteries",
+        author: "Cuttlefish That Loves Diving",
+        description: null,
+        isbn: null,
+        isbn13: null,
+        identifiersJson: JSON.stringify({ isbn10: null, isbn13: null, candidates: [] }),
+        localMetadataJson: JSON.stringify({
+          title: "Lord of the Mysteries",
+          author: "Cuttlefish That Loves Diving",
+          year: null,
+          isbn: null,
+          fileTitle: "Lord of the Mysteries",
+          raw: { title: "Lord of the Mysteries", author: "Cuttlefish That Loves Diving", publishedDate: null, fileName: "Lord of the Mysteries" }
+        }),
+        language: null,
+        publisher: null,
+        publishedDate: "2018",
+        format: "epub",
+        filePath: "/library/Lord of the Mysteries.epub",
+        coverPath: null,
+        metadataStatus: "partial",
+        metadataProvenanceJson: JSON.stringify({ publishedDate: "existing" }),
+        scanStatus: "succeeded",
+        enrichmentStatus: "succeeded",
+        fileSize: 42,
+        fileMtime: now,
+        createdAt: now,
+        updatedAt: now
+      })
+      .returning()
+      .get();
+    const openLibraryProvider: MetadataProvider = {
+      id: "openlibrary",
+      lookup: async () => null,
+      searchByMetadata: async () => [
+        {
+          title: "Lord of the Mysteries",
+          author: "Cuttlefish That Loves Diving",
+          publishedDate: "0101-01-01T00:00:00+00:00",
+          description: "A mysterious sequence begins."
+        }
+      ]
+    };
+    const service = new MetadataService(
+      database,
+      providerStub("local") as never,
+      openLibraryProvider as never,
+      providerStub("googlebooks") as never,
+      providerStub("bnf") as never,
+      providerStub("mangadex") as never,
+      providerStub("comicvine") as never
+    );
+
+    const result = await service.enrichBook(book.id);
+
+    expect(result.publishedDate).toBe("2018");
+
+    database.onModuleDestroy();
+  });
+
+  it.runIf(canLoadBetterSqlite())("records MangaDex provenance and stores its cover locally", async () => {
+    const database = createDatabase();
+    const fetchMock = vi.spyOn(globalThis, "fetch").mockResolvedValue({
+      ok: true,
+      headers: new Headers({ "content-type": "image/jpeg" }),
+      arrayBuffer: async () => new Uint8Array([9, 8, 7]).buffer
+    } as Response);
+    const mangaDexProvider: MetadataProvider = {
+      id: "mangadex",
+      lookup: async () => null,
+      searchByMetadata: async () => [
+        {
+          title: "Solo Leveling",
+          author: "Chugong",
+          description: "A hunter levels up alone.",
+          publishedDate: "2018",
+          coverUrl: "https://uploads.mangadex.org/covers/manga-1/cover.jpg.512.jpg"
+        }
+      ]
+    };
+    const service = new MetadataService(
+      database,
+      providerStub("local") as never,
+      providerStub("openlibrary") as never,
+      providerStub("googlebooks") as never,
+      providerStub("bnf") as never,
+      mangaDexProvider as never,
+      providerStub("comicvine") as never
+    );
+
+    const result = await service.enrichMetadata(
+      {
+        title: "Solo Leveling T03",
+        author: null,
+        description: null,
+        isbn: null,
+        language: null,
+        publisher: null,
+        publishedDate: null,
+        coverPath: null
+      },
+      "/library/Solo Leveling T03.cbz",
+      { remote: true }
+    );
+
+    expect(fetchMock).toHaveBeenCalledWith("https://uploads.mangadex.org/covers/manga-1/cover.jpg.512.jpg", expect.any(Object));
+    expect(JSON.parse(result.metadataProvenanceJson)).toMatchObject({
+      author: "mangadex",
+      description: "mangadex",
+      publishedDate: "mangadex",
+      coverPath: "mangadex"
+    });
+    expect(result.coverPath).toMatch(/storage\/covers\/.+\.jpg$/);
+
+    database.onModuleDestroy();
+  });
+
+  it.runIf(canLoadBetterSqlite())("scores MangaDex matches against the cleaned series title instead of the noisy archive title", async () => {
+    const database = createDatabase();
+    const mangaDexProvider: MetadataProvider = {
+      id: "mangadex",
+      lookup: async () => null,
+      searchByMetadata: async () => [
+        {
+          title: "Dragon Ball SD",
+          author: "Naho Ooishi",
+          description: "A super-deformed Dragon Ball spin-off.",
+          publishedDate: "2010",
+          sourceId: "dragon-ball-sd"
+        }
+      ]
+    };
+    const service = new MetadataService(
+      database,
+      providerStub("local") as never,
+      providerStub("openlibrary") as never,
+      providerStub("googlebooks") as never,
+      providerStub("bnf") as never,
+      mangaDexProvider as never,
+      providerStub("comicvine") as never
+    );
+
+    const result = await service.enrichMetadata(
+      {
+        title: "Dragon.Ball.SD.T01.FRENCH.CBZ.eBook-Paprika+",
+        author: null,
+        description: null,
+        isbn: null,
+        language: null,
+        publisher: null,
+        publishedDate: null,
+        coverPath: null
+      },
+      "/library/Dragon.Ball.SD.T01.FRENCH.CBZ.eBook-Paprika+.cbz",
+      { remote: true }
+    );
+
+    expect(result).toMatchObject({
+      title: "Dragon Ball SD",
+      author: "Naho Ooishi",
+      description: "A super-deformed Dragon Ball spin-off.",
+      publishedDate: "2010"
+    });
+    expect(JSON.parse(result.metadataProvenanceJson)).toMatchObject({
+      title: "local",
+      author: "mangadex",
+      description: "mangadex"
+    });
+
+    database.onModuleDestroy();
+  });
+
+  it.runIf(canLoadBetterSqlite())("accepts MangaDex alias matches through the provider score title", async () => {
+    const database = createDatabase();
+    const mangaDexProvider: MetadataProvider = {
+      id: "mangadex",
+      lookup: async () => null,
+      searchByMetadata: async () => [
+        {
+          title: "Demon Slayer: Kimetsu Academy",
+          scoreTitle: "Demon Slayer Kimetsu Academy",
+          author: "Natsuki Hokami",
+          description: "School spin-off.",
+          publishedDate: "2021",
+          sourceId: "kimetsu-academy"
+        }
+      ]
+    };
+    const service = new MetadataService(
+      database,
+      providerStub("local") as never,
+      providerStub("openlibrary") as never,
+      providerStub("googlebooks") as never,
+      providerStub("bnf") as never,
+      mangaDexProvider as never,
+      providerStub("comicvine") as never
+    );
+
+    const result = await service.enrichMetadata(
+      {
+        title: "Demon.Slayer.School.Days.T01.FRENCH.CBZ. -ebdz",
+        author: null,
+        description: null,
+        isbn: null,
+        language: null,
+        publisher: null,
+        publishedDate: null,
+        coverPath: null
+      },
+      "/library/Demon.Slayer.School.Days.T01.FRENCH.CBZ. -ebdz.cbz",
+      { remote: true }
+    );
+
+    expect(result).toMatchObject({
+      title: "Demon Slayer School Days",
+      author: "Natsuki Hokami",
+      description: "School spin-off.",
+      publishedDate: "2021"
+    });
+
+    database.onModuleDestroy();
+  });
+
+  it.runIf(canLoadBetterSqlite())("re-enriches existing manga with the cleaned series title on the live book path", async () => {
+    const database = createDatabase();
+    mkdirSync(database.config.storageDir, { recursive: true });
+    const filePath = join(database.config.storageDir, "Demon.Slayer.School.Days.T01.FRENCH.CBZ. -ebdz.cbz");
+    const zip = new AdmZip();
+    zip.addFile("001.jpg", Buffer.from([0xff, 0xd8, 0xff, 0xd9]));
+    zip.writeZip(filePath);
+    const now = database.now();
+    const library = database.db
+      .insert(libraries)
+      .values({ name: "Manga", path: database.config.storageDir, enabled: true, createdAt: now, updatedAt: now })
+      .returning()
+      .get();
+    const book = database.db
+      .insert(books)
+      .values({
+        libraryId: library.id,
+        title: "Demon.Slayer.School.Days.T01.FRENCH.CBZ. -ebdz",
+        author: null,
+        description: null,
+        isbn: null,
+        isbn13: null,
+        identifiersJson: JSON.stringify({ isbn10: null, isbn13: null, candidates: [] }),
+        localMetadataJson: JSON.stringify({
+          title: "Demon Slayer School Days T01 FRENCH",
+          author: null,
+          year: null,
+          isbn: null,
+          fileTitle: "Demon Slayer School Days T01 FRENCH",
+          raw: {
+            title: "Demon.Slayer.School.Days.T01.FRENCH.CBZ. -ebdz",
+            author: null,
+            publishedDate: null,
+            fileName: "Demon.Slayer.School.Days.T01.FRENCH.CBZ. -ebdz"
+          }
+        }),
+        language: null,
+        publisher: null,
+        publishedDate: null,
+        format: "cbz",
+        filePath,
+        coverPath: null,
+        metadataStatus: "none",
+        metadataProvenanceJson: JSON.stringify({ title: "local" }),
+        scanStatus: "succeeded",
+        enrichmentStatus: "succeeded",
+        fileSize: 42,
+        fileMtime: now,
+        createdAt: now,
+        updatedAt: now
+      })
+      .returning()
+      .get();
+    const mangaDexSearch = vi.fn<(_: MetadataSearchQuery, __: MetadataProviderConfig) => Promise<MetadataMatch[]>>(async () => [
+      {
+        title: "Demon Slayer School Days",
+        author: "Natsuki Hokami",
+        description: "School spin-off.",
+        publishedDate: "2021",
+        sourceId: "demon-slayer-school-days"
+      }
+    ]);
+    const service = new MetadataService(
+      database,
+      providerStub("local") as never,
+      providerStub("openlibrary") as never,
+      providerStub("googlebooks") as never,
+      providerStub("bnf") as never,
+      { id: "mangadex", lookup: async () => null, searchByMetadata: mangaDexSearch } as never,
+      providerStub("comicvine") as never
+    );
+
+    const result = await service.enrichBook(book.id);
+
+    expect(mangaDexSearch.mock.calls[0]?.[0].title).toBe("Demon Slayer School Days");
+    expect(result).toMatchObject({
+      title: "Demon Slayer School Days",
+      author: "Natsuki Hokami",
+      description: "School spin-off.",
+      publishedDate: "2021"
+    });
+
+    database.onModuleDestroy();
+  });
 });
 
 function createDatabase(): DatabaseService {
diff --git a/apps/api/src/metadata/metadata.service.ts b/apps/api/src/metadata/metadata.service.ts
index 1553ba0..005ca2a 100644
--- a/apps/api/src/metadata/metadata.service.ts
+++ b/apps/api/src/metadata/metadata.service.ts
@@ -1,21 +1,45 @@
 import { BadRequestException, Injectable, NotFoundException } from "@nestjs/common";
+import { createHash } from "node:crypto";
+import { existsSync, mkdirSync, writeFileSync } from "node:fs";
+import { dirname, extname, join } from "node:path";
 import { eq } from "drizzle-orm";
 import {
   MetadataSourcesConfigDto,
   UpdateMetadataSourcesConfigDto
 } from "@readabook/shared";
 import { DatabaseService } from "../database/database.service.js";
-import { automationSettings, books, metadataSourceConfig } from "../database/schema.js";
-import { BookMetadata } from "../scanner/metadata.js";
+import { automationSettings, books, metadataSourceConfig, series } from "../database/schema.js";
+import { BookMetadata, extractMetadata } from "../scanner/metadata.js";
 import { BnfProvider } from "./adapters/bnf.provider.js";
+import { ComicVineProvider } from "./adapters/comic-vine.provider.js";
 import { GoogleBooksProvider } from "./adapters/google-books.provider.js";
 import { LocalMetadataProvider } from "./adapters/local.provider.js";
+import { MangaDexProvider } from "./adapters/mangadex.provider.js";
 import { OpenLibraryProvider } from "./adapters/open-library.provider.js";
-import { BookIdentifiers, LocalMetadataHints, MetadataMatch, MetadataProvider, MetadataProviderConfig } from "./metadata.types.js";
+import { describeMetadataProviderError, providerFetch } from "./adapters/provider-fetch.js";
+import {
+  BookIdentifiers,
+  LocalMetadataHints,
+  MetadataField,
+  MetadataMatch,
+  MetadataProvider,
+  MetadataProviderConfig,
+  MetadataProviderId,
+  MetadataProvenance,
+  MetadataSearchQuery,
+  MetadataStatus
+} from "./metadata.types.js";
 import { ExtractIdentifiers, toIsbn13 } from "./use-cases/extract-identifiers.js";
 import { ExtractLocalMetadataHints } from "./use-cases/extract-local-metadata-hints.js";
+import { extractSeriesVolume } from "./use-cases/extract-series-volume.js";
+import { normalizePublishedDate } from "./use-cases/normalize-published-date.js";
 import { ResolveProviderChain } from "./use-cases/resolve-provider-chain.js";
-import { ScoreMetadataMatch } from "./use-cases/score-metadata-match.js";
+import { ScoredMetadataMatch, ScoreMetadataMatch } from "./use-cases/score-metadata-match.js";
+
+type ProviderCandidate = ScoredMetadataMatch & {
+  provider: MetadataProviderId;
+  priority: number;
+};
 
 @Injectable()
 export class MetadataService {
@@ -29,9 +53,11 @@ export class MetadataService {
     local: LocalMetadataProvider,
     openLibrary: OpenLibraryProvider,
     googleBooks: GoogleBooksProvider,
-    bnf: BnfProvider
+    bnf: BnfProvider,
+    mangaDex: MangaDexProvider,
+    comicVine: ComicVineProvider
   ) {
-    this.resolveProviderChain = new ResolveProviderChain([local, openLibrary, googleBooks, bnf]);
+    this.resolveProviderChain = new ResolveProviderChain([local, openLibrary, googleBooks, bnf, mangaDex, comicVine]);
   }
 
   getSourcesConfig(): MetadataSourcesConfigDto {
@@ -75,96 +101,119 @@ export class MetadataService {
     localMetadata: BookMetadata,
     filePath: string,
     options: { remote: boolean }
-  ): Promise<BookMetadata & { isbn13: string | null; identifiersJson: string; localMetadataJson: string }> {
+  ): Promise<
+    BookMetadata & {
+      isbn13: string | null;
+      identifiersJson: string;
+      localMetadataJson: string;
+      metadataStatus: MetadataStatus;
+      metadataProvenanceJson: string;
+    }
+  > {
     const identifiers = this.extractIdentifiers.fromMetadataAndFile(localMetadata, filePath);
     const local = this.extractLocalMetadataHints.fromMetadataAndFile(localMetadata, filePath);
     const configs = this.getProviderConfigs();
     const chain = options.remote
       ? this.resolveProviderChain.resolve(configs)
       : this.resolveProviderChain.resolve(configs).filter((entry) => entry.provider.id === "local");
-    let merged: BookMetadata = { ...localMetadata };
+    const candidates: ProviderCandidate[] = [];
+    const query = this.buildSearchQuery(local, identifiers, filePath);
 
     for (const { provider, config } of chain) {
+      if (provider.id === "local") continue;
       try {
         const hasIsbn = Boolean(identifiers.isbn13 ?? identifiers.isbn10);
-        const match =
-          provider.id === "local" || hasIsbn
-            ? await provider.lookup(
-                {
-                  title: merged.title,
-                  author: merged.author,
-                  filePath,
-                  sourceId: null,
-                  identifiers,
-                  local
-                },
-                config
-              )
-            : null;
+        const match = hasIsbn
+          ? await provider.lookup(
+              {
+                title: local.title,
+                author: local.author,
+                filePath,
+                sourceId: null,
+                identifiers,
+                local
+              },
+              config
+            )
+          : null;
         if (match) {
-          merged = mergeMetadata(merged, match);
+          const completedMatch = match.description
+            ? match
+            : mergeMetadataMatch(match, await this.searchMissingDescription(provider, config, local, identifiers, filePath));
+          candidates.push(scoreProviderCandidate(this.scoreMetadataMatch, query, completedMatch, provider.id, config.priority));
           continue;
         }
-        if (!options.remote || provider.id === "local") continue;
-        const query = {
-          title: local.title,
-          author: local.author,
-          year: local.year,
-          isbn: identifiers.isbn13 ?? identifiers.isbn10 ?? local.isbn
-        };
-        const best = this.scoreMetadataMatch.best(query, await provider.searchByMetadata(query, config));
+        if (!options.remote) continue;
+        const matches = await provider.searchByMetadata(query, config);
+        const best = this.scoreMetadataMatch.best(query, matches);
+        if (!best && matches.length) {
+          console.info(
+            `[metadata] Provider ${provider.id} returned ${matches.length} result(s) rejected by scoring for "${query.title}"`
+          );
+        }
         if (best) {
           if (!isActionableSearchMatch(best.match)) continue;
-          merged = shouldReplaceAmbiguousIdentification(localMetadata, merged, best.match)
-            ? mergeMetadata({ ...localMetadata, author: null, isbn: null, description: null, language: null, publisher: null, publishedDate: null }, best.match)
-            : mergeMetadata(merged, best.match);
+          let providerMatch = best.match;
 
-          const detailedMatch = await this.lookupSearchMatchDetails(provider, config, filePath, identifiers, local, merged, best.match);
-          if (detailedMatch) merged = mergeMetadata(merged, detailedMatch);
+          const detailedMatch = await this.lookupSearchMatchDetails(provider, config, filePath, identifiers, local, best.match);
+          if (detailedMatch) providerMatch = mergeMetadataMatch(detailedMatch, best.match);
+          candidates.push(scoreProviderCandidate(this.scoreMetadataMatch, query, providerMatch, provider.id, config.priority));
         }
-      } catch {
-        // Provider failures must not block local ingestion.
+      } catch (error) {
+        console.warn(`[metadata] Provider ${provider.id} failed for "${query.title}": ${describeMetadataProviderError(error)}`);
       }
     }
 
+    const materializedCandidates = await this.materializeQualifiedCovers(candidates, filePath);
+    const localProvenance = provenanceFromLocal(localMetadata);
+    const { metadata: merged, provenance: remoteProvenance } = mergeCandidatesWithLocal(
+      materializedCandidates,
+      localMetadata,
+      identifiers,
+      query.title
+    );
+    let provenance: MetadataProvenance = { ...localProvenance, ...remoteProvenance };
+    if (merged.isbn && !provenance.isbn) provenance.isbn = "local";
     const isbn13 = identifiers.isbn13 ?? (merged.isbn ? toIsbn13(merged.isbn) : null);
+    const metadataStatus = computeMetadataStatus(merged);
     return {
       ...merged,
       isbn: merged.isbn ?? isbn13 ?? identifiers.isbn10,
       isbn13,
       identifiersJson: JSON.stringify(identifiers),
-      localMetadataJson: JSON.stringify(local)
+      localMetadataJson: JSON.stringify(local),
+      metadataStatus,
+      metadataProvenanceJson: JSON.stringify(provenance)
     };
   }
 
   async enrichBook(bookId: number): Promise<typeof books.$inferSelect> {
     const book = this.database.db.select().from(books).where(eq(books.id, bookId)).get();
     if (!book) throw new NotFoundException("Book not found");
-    const metadata: BookMetadata = {
-      title: book.title,
-      author: book.author,
-      description: book.description,
-      isbn: book.isbn,
-      language: book.language,
-      publisher: book.publisher,
-      publishedDate: book.publishedDate,
-      coverPath: book.coverPath
-    };
+    const local = parseStoredLocalMetadata(book.localMetadataJson);
+    const metadata = await this.extractCurrentLocalMetadata(book, local);
     const enriched = await this.enrichMetadata(metadata, book.filePath, { remote: true });
+    const next = preserveExistingWhenMissing(enriched, book, book.filePath);
+    const seriesInfo = this.resolveSeries(next.title, book.filePath);
     return this.database.db
       .update(books)
       .set({
-        title: enriched.title,
-        author: enriched.author,
-        description: enriched.description,
-        isbn: enriched.isbn,
-        isbn13: enriched.isbn13,
+        seriesId: seriesInfo.seriesId,
+        title: next.title,
+        author: next.author,
+        description: next.description,
+        isbn: next.isbn,
+        isbn13: next.isbn13,
         identifiersJson: enriched.identifiersJson,
         localMetadataJson: enriched.localMetadataJson,
-        language: enriched.language,
-        publisher: enriched.publisher,
-        publishedDate: enriched.publishedDate,
-        coverPath: enriched.coverPath,
+        language: next.language,
+        publisher: next.publisher,
+        publishedDate: next.publishedDate,
+        volumeNumber: seriesInfo.volumeNumber,
+        volumeLabel: seriesInfo.volumeLabel,
+        coverPath: next.coverPath,
+        metadataStatus: next.metadataStatus,
+        metadataProvenanceJson: next.metadataProvenanceJson,
         updatedAt: this.database.now()
       })
       .where(eq(books.id, book.id))
@@ -172,6 +221,57 @@ export class MetadataService {
       .get();
   }
 
+  private async extractCurrentLocalMetadata(book: typeof books.$inferSelect, local: LocalMetadataHints | null): Promise<BookMetadata> {
+    const fallback: BookMetadata = {
+      title: local?.title ?? book.title,
+      author: local ? local.author : book.author,
+      description: null,
+      isbn: local ? local.isbn : book.isbn,
+      language: null,
+      publisher: book.publisher,
+      publishedDate: normalizePublishedDate(local ? local.year : book.publishedDate),
+      coverPath: book.coverPath
+    };
+    if (!existsSync(book.filePath)) return fallback;
+    try {
+      const extracted = await extractMetadata(book.filePath, this.database.config.storageDir);
+      return {
+        title: extracted.title || fallback.title,
+        author: extracted.author ?? fallback.author,
+        description: extracted.description ?? fallback.description,
+        isbn: extracted.isbn ?? fallback.isbn,
+        language: extracted.language ?? fallback.language,
+        publisher: extracted.publisher ?? fallback.publisher,
+        publishedDate: normalizePublishedDate(extracted.publishedDate) ?? fallback.publishedDate,
+        coverPath: extracted.coverPath ?? fallback.coverPath
+      };
+    } catch {
+      return fallback;
+    }
+  }
+
+  private resolveSeries(title: string, filePath: string): { seriesId: number; volumeNumber: number | null; volumeLabel: string | null } {
+    const parsed = extractSeriesVolume(title, filePath);
+    const now = this.database.now();
+    const row = this.database.db
+      .insert(series)
+      .values({
+        title: parsed.seriesTitle,
+        normalizedTitle: parsed.normalizedSeriesTitle,
+        description: null,
+        publisher: null,
+        createdAt: now,
+        updatedAt: now
+      })
+      .onConflictDoUpdate({
+        target: series.normalizedTitle,
+        set: { title: parsed.seriesTitle, updatedAt: now }
+      })
+      .returning({ id: series.id })
+      .get();
+    return { seriesId: row.id, volumeNumber: parsed.volumeNumber, volumeLabel: parsed.volumeLabel };
+  }
+
   private getProviderConfigs(): MetadataProviderConfig[] {
     return this.database.db
       .select()
@@ -195,7 +295,6 @@ export class MetadataService {
     filePath: string,
     identifiers: BookIdentifiers,
     local: LocalMetadataHints,
-    merged: BookMetadata,
     match: MetadataMatch
   ): Promise<MetadataMatch | null> {
     const derivedIdentifiers = {
@@ -206,13 +305,12 @@ export class MetadataService {
     };
     const hasNewIdentifier = derivedIdentifiers.isbn13 !== identifiers.isbn13 || derivedIdentifiers.isbn10 !== identifiers.isbn10;
     const hasLookupTarget = hasNewIdentifier || Boolean(match.sourceId);
-    const needsDetails = !merged.description && hasLookupTarget;
-    if (!hasLookupTarget && !needsDetails) return null;
+    if (!hasLookupTarget) return null;
 
     return provider.lookup(
       {
-        title: merged.title,
-        author: merged.author,
+        title: match.title ?? local.title,
+        author: match.author ?? local.author,
         filePath,
         sourceId: match.sourceId,
         identifiers: derivedIdentifiers,
@@ -221,26 +319,309 @@ export class MetadataService {
       config
     );
   }
+
+  private async searchMissingDescription(
+    provider: MetadataProvider,
+    config: MetadataProviderConfig,
+    local: LocalMetadataHints,
+    identifiers: BookIdentifiers,
+    filePath: string
+  ): Promise<MetadataMatch | null> {
+    const query = this.buildSearchQuery(local, identifiers, filePath);
+    const matches = await provider.searchByMetadata(query, config);
+    return (
+      matches
+        .map((match) => ({ match, score: this.scoreMetadataMatch.score(query, match) }))
+        .filter((entry) => entry.match.description && entry.score >= 75)
+        .sort((left, right) => right.score - left.score)[0]?.match ?? null
+    );
+  }
+
+  private buildSearchQuery(local: LocalMetadataHints, identifiers: BookIdentifiers, filePath: string): MetadataSearchQuery {
+    return {
+      title: extractSeriesVolume(local.title, filePath).seriesTitle,
+      author: local.author,
+      year: local.year,
+      isbn: identifiers.isbn13 ?? identifiers.isbn10 ?? local.isbn
+    };
+  }
+
+  private async materializeCover(match: MetadataMatch, filePath: string, provider: MetadataProviderId): Promise<MetadataMatch> {
+    if (match.coverPath || !match.coverUrl) return match;
+    try {
+      const response = await providerFetch("cover", match.coverUrl, { timeoutMs: 5000 });
+      if (!response.ok) return match;
+      const data = Buffer.from(await response.arrayBuffer());
+      if (!data.length) return match;
+      const extension = coverExtension(match.coverUrl, response.headers.get("content-type"));
+      const hash = createHash("sha256").update(`${filePath}:${provider}:${match.coverUrl}`).digest("hex").slice(0, 24);
+      const target = join(this.database.config.storageDir, "covers", `${hash}${extension}`);
+      mkdirSync(dirname(target), { recursive: true });
+      writeFileSync(target, data);
+      return { ...match, coverPath: target };
+    } catch {
+      return match;
+    }
+  }
+
+  private async materializeQualifiedCovers(candidates: ProviderCandidate[], filePath: string): Promise<ProviderCandidate[]> {
+    const materialized: ProviderCandidate[] = [];
+    for (const candidate of candidates) {
+      if (isQualifiedSourceCover(candidate)) {
+        materialized.push({
+          ...candidate,
+          match: await this.materializeCover(candidate.match, filePath, candidate.provider)
+        });
+      } else {
+        materialized.push(candidate);
+      }
+    }
+    return materialized;
+  }
 }
 
 function isActionableSearchMatch(match: MetadataMatch): boolean {
   return Boolean(match.isbn ?? match.sourceId ?? match.description);
 }
 
-function shouldReplaceAmbiguousIdentification(local: BookMetadata, current: BookMetadata, next: MetadataMatch): boolean {
-  if (local.author || local.isbn || !current.isbn || !next.isbn || current.isbn === next.isbn) return false;
-  return Boolean(next.description);
-}
-
-function mergeMetadata(current: BookMetadata, next: MetadataMatch): BookMetadata {
+function mergeMetadataMatch(current: MetadataMatch, next: MetadataMatch | null): MetadataMatch {
+  if (!next) return current;
   return {
-    title: next.title ?? current.title,
-    author: current.author ?? next.author ?? null,
-    description: current.description ?? next.description ?? null,
-    isbn: current.isbn ?? next.isbn ?? null,
-    language: current.language ?? next.language ?? null,
-    publisher: current.publisher ?? next.publisher ?? null,
-    publishedDate: current.publishedDate ?? next.publishedDate ?? null,
-    coverPath: current.coverPath ?? next.coverPath ?? null
+    title: current.title ?? next.title,
+    author: current.author ?? next.author,
+    description: current.description ?? next.description,
+    isbn: current.isbn ?? next.isbn,
+    language: current.language ?? next.language,
+    publisher: current.publisher ?? next.publisher,
+    publishedDate: normalizePublishedDate(current.publishedDate) ?? normalizePublishedDate(next.publishedDate),
+    coverPath: current.coverPath ?? next.coverPath,
+    coverUrl: current.coverUrl ?? next.coverUrl,
+    sourceId: current.sourceId ?? next.sourceId,
+    identifiers: current.identifiers ?? next.identifiers
   };
 }
+
+const metadataFields: MetadataField[] = ["title", "author", "description", "isbn", "language", "publisher", "publishedDate", "coverPath"];
+const fillableMetadataFields: MetadataField[] = ["author", "description", "isbn", "language", "publisher", "publishedDate"];
+
+function scoreProviderCandidate(
+  scorer: ScoreMetadataMatch,
+  query: MetadataSearchQuery,
+  match: MetadataMatch,
+  provider: MetadataProviderId,
+  priority: number
+): ProviderCandidate {
+  return { ...scorer.details(query, match), provider, priority };
+}
+
+function mergeCandidatesWithLocal(
+  candidates: ProviderCandidate[],
+  local: BookMetadata,
+  identifiers: BookIdentifiers,
+  title: string
+): { metadata: BookMetadata; provenance: MetadataProvenance } {
+  const sorted = [...candidates].sort(compareProviderCandidates);
+  const retained = sorted[0] ?? null;
+  const completionOrder = [...(retained ? [retained] : []), ...sorted.filter((candidate) => candidate !== retained)];
+  const metadata: BookMetadata = {
+    title,
+    author: local.author ?? null,
+    description: local.description ?? null,
+    isbn: identifiers.isbn13 ?? identifiers.isbn10 ?? local.isbn ?? null,
+    language: local.language ?? null,
+    publisher: local.publisher ?? null,
+    publishedDate: normalizePublishedDate(local.publishedDate),
+    coverPath: local.coverPath ?? null
+  };
+  const provenance: MetadataProvenance = {};
+  for (const field of fillableMetadataFields) {
+    if (hasMetadataValue(metadata[field])) continue;
+    const source = completionOrder.find((candidate) => hasMetadataValue(normalizeCandidateField(candidate.match, field)));
+    if (!source) continue;
+    metadata[field] = normalizeCandidateField(source.match, field) as never;
+    provenance[field] = source.provider;
+    provenance[`${field}Score` as MetadataField] = String(source.score) as never;
+  }
+  const coverSource = completionOrder.find((candidate) => isQualifiedSourceCover(candidate) && hasMetadataValue(candidate.match.coverPath));
+  if (coverSource && shouldUseSourceCover(metadata.coverPath)) {
+    metadata.coverPath = coverSource.match.coverPath ?? null;
+    provenance.coverPath = coverSource.provider;
+    provenance.coverPathScore = String(coverSource.score) as never;
+  }
+  return { metadata, provenance };
+}
+
+function compareProviderCandidates(left: ProviderCandidate, right: ProviderCandidate): number {
+  if (left.isbnMatch !== right.isbnMatch) return left.isbnMatch ? -1 : 1;
+  const leftHighConfidence = isHighConfidenceSelection(left);
+  const rightHighConfidence = isHighConfidenceSelection(right);
+  if (leftHighConfidence && rightHighConfidence) return left.priority - right.priority;
+  if (leftHighConfidence !== rightHighConfidence) return leftHighConfidence ? -1 : 1;
+  if (left.score !== right.score) return right.score - left.score;
+  return left.priority - right.priority;
+}
+
+function isHighConfidenceSelection(candidate: ProviderCandidate): boolean {
+  return candidate.titleScore > 90 && (candidate.authorScore == null || candidate.authorScore >= 15);
+}
+
+function isQualifiedSourceCover(candidate: ProviderCandidate): boolean {
+  return candidate.score >= 80 && candidate.titleScore >= 85 && Boolean(candidate.match.coverPath ?? candidate.match.coverUrl);
+}
+
+function shouldUseSourceCover(currentCoverPath: string | null): boolean {
+  return !hasMetadataValue(currentCoverPath) || isLocalCoverPath(currentCoverPath);
+}
+
+function isLocalCoverPath(value: string): boolean {
+  return /[/\\]covers[/\\][a-f0-9]{24}\.[a-z0-9]+$/i.test(value);
+}
+
+function normalizeCandidateField(match: MetadataMatch, field: MetadataField): string | null {
+  if (field === "publishedDate") return normalizePublishedDate(match.publishedDate);
+  return match[field] ?? null;
+}
+
+function hasMetadataValue(value: string | null | undefined): value is string {
+  return Boolean(value && value.trim());
+}
+
+function provenanceFromLocal(local: BookMetadata): MetadataProvenance {
+  const provenance: MetadataProvenance = {};
+  for (const field of metadataFields) {
+    if (local[field] != null && local[field] !== "") provenance[field] = "local";
+  }
+  return provenance;
+}
+
+function computeMetadataStatus(metadata: Pick<BookMetadata, "author" | "description" | "isbn" | "language" | "publisher" | "publishedDate" | "coverPath">): MetadataStatus {
+  const hasCover = Boolean(metadata.coverPath);
+  const filled = [metadata.author, metadata.description, metadata.isbn, metadata.language, metadata.publisher, metadata.publishedDate].filter(Boolean).length;
+  if (hasCover && filled >= 2) return "enriched";
+  if (hasCover || filled > 0) return "partial";
+  return "none";
+}
+
+function mergeExistingProvenance(
+  enriched: BookMetadata & { metadataProvenanceJson: string },
+  existing: typeof books.$inferSelect,
+  finalValues: BookMetadata & { isbn13: string | null }
+): MetadataProvenance {
+  const next = parseProvenance(enriched.metadataProvenanceJson);
+  const previous = parseProvenance(existing.metadataProvenanceJson);
+  const provenance: MetadataProvenance = { ...previous };
+  for (const field of metadataFields) {
+    if (field === "title") {
+      if (finalValues.title && !provenance.title) {
+        provenance.title = finalValues.title === enriched.title && finalValues.title !== existing.title ? (next.title ?? "local") : (previous.title ?? "existing");
+      }
+      continue;
+    }
+    if (field === "publishedDate") {
+      const finalDate = normalizePublishedDate(finalValues.publishedDate);
+      if (finalDate && finalDate === normalizePublishedDate(enriched.publishedDate) && finalDate !== normalizePublishedDate(existing.publishedDate)) {
+        provenance[field] = next[field] ?? provenance[field];
+        copyScoreProvenance(next, provenance, field);
+      } else if (finalDate && !provenance[field]) {
+        provenance[field] = previous[field] ?? "existing";
+      }
+      continue;
+    }
+    if (field === "coverPath" && finalValues.coverPath && finalValues.coverPath === enriched.coverPath && finalValues.coverPath !== existing.coverPath) {
+      provenance.coverPath = next.coverPath ?? provenance.coverPath;
+      copyScoreProvenance(next, provenance, field);
+      continue;
+    }
+    if (finalValues[field] && finalValues[field] === enriched[field] && finalValues[field] !== existing[field]) {
+      provenance[field] = next[field] ?? provenance[field];
+      copyScoreProvenance(next, provenance, field);
+      continue;
+    }
+    if (finalValues[field] != null && existing[field] != null && !provenance[field]) {
+      provenance[field] = previous[field] ?? "existing";
+    }
+  }
+  return provenance;
+}
+
+function copyScoreProvenance(source: MetadataProvenance, target: MetadataProvenance, field: MetadataField): void {
+  const scoreKey = `${field}Score`;
+  if (source[scoreKey]) target[scoreKey] = source[scoreKey];
+}
+
+function parseProvenance(value: string | null): MetadataProvenance {
+  if (!value) return {};
+  try {
+    const parsed = JSON.parse(value) as MetadataProvenance;
+    return parsed && typeof parsed === "object" ? parsed : {};
+  } catch {
+    return {};
+  }
+}
+
+function coverExtension(url: string, contentType: string | null): string {
+  if (contentType?.includes("png")) return ".png";
+  if (contentType?.includes("webp")) return ".webp";
+  if (contentType?.includes("gif")) return ".gif";
+  const fromUrl = extname(new URL(url).pathname).toLowerCase();
+  return fromUrl === ".png" || fromUrl === ".webp" || fromUrl === ".gif" || fromUrl === ".jpg" || fromUrl === ".jpeg" ? fromUrl : ".jpg";
+}
+
+function parseStoredLocalMetadata(value: string | null): LocalMetadataHints | null {
+  if (!value) return null;
+  try {
+    const parsed = JSON.parse(value) as Partial<LocalMetadataHints>;
+    return typeof parsed.title === "string" ? (parsed as LocalMetadataHints) : null;
+  } catch {
+    return null;
+  }
+}
+
+function preserveExistingWhenMissing(
+  enriched: BookMetadata & {
+    isbn13: string | null;
+    identifiersJson: string;
+    localMetadataJson: string;
+    metadataStatus: MetadataStatus;
+    metadataProvenanceJson: string;
+  },
+  existing: typeof books.$inferSelect,
+  filePath: string
+): BookMetadata & { isbn13: string | null; metadataStatus: MetadataStatus; metadataProvenanceJson: string } {
+  const previousProvenance = parseProvenance(existing.metadataProvenanceJson);
+  const next = {
+    title: chooseTitle(existing.title, enriched.title, filePath),
+    author: existing.author ?? enriched.author,
+    description: existing.description ?? enriched.description,
+    isbn: existing.isbn ?? enriched.isbn,
+    isbn13: existing.isbn13 ?? enriched.isbn13,
+    language: existing.language ?? enriched.language,
+    publisher: existing.publisher ?? enriched.publisher,
+    publishedDate: normalizePublishedDate(existing.publishedDate) ?? normalizePublishedDate(enriched.publishedDate),
+    coverPath: chooseCoverPath(existing.coverPath, enriched.coverPath, previousProvenance)
+  };
+  const provenance = mergeExistingProvenance(enriched, existing, next);
+  return {
+    ...next,
+    metadataStatus: computeMetadataStatus(next),
+    metadataProvenanceJson: JSON.stringify(provenance)
+  };
+}
+
+function chooseCoverPath(existingCoverPath: string | null, enrichedCoverPath: string | null, previousProvenance: MetadataProvenance): string | null {
+  if (!existingCoverPath) return enrichedCoverPath;
+  if (!enrichedCoverPath || enrichedCoverPath === existingCoverPath) return existingCoverPath;
+  return canReplaceExistingCover(existingCoverPath, previousProvenance) ? enrichedCoverPath : existingCoverPath;
+}
+
+function chooseTitle(existingTitle: string, enrichedTitle: string, filePath: string): string {
+  if (!enrichedTitle) return existingTitle;
+  if (!existingTitle) return enrichedTitle;
+  const parsedExisting = extractSeriesVolume(existingTitle, filePath).seriesTitle;
+  return parsedExisting === enrichedTitle && existingTitle !== enrichedTitle ? enrichedTitle : existingTitle;
+}
+
+function canReplaceExistingCover(existingCoverPath: string, previousProvenance: MetadataProvenance): boolean {
+  const provenance = previousProvenance.coverPath;
+  return (provenance == null || provenance === "local" || provenance === "existing") && isLocalCoverPath(existingCoverPath);
+}
diff --git a/apps/api/src/metadata/metadata.types.ts b/apps/api/src/metadata/metadata.types.ts
index c23b836..f2200da 100644
--- a/apps/api/src/metadata/metadata.types.ts
+++ b/apps/api/src/metadata/metadata.types.ts
@@ -1,6 +1,6 @@
 import { BookMetadata } from "../scanner/metadata.js";
 
-export type MetadataProviderId = "local" | "openlibrary" | "googlebooks" | "bnf";
+export type MetadataProviderId = "local" | "openlibrary" | "googlebooks" | "bnf" | "mangadex" | "comicvine";
 
 export type BookIdentifiers = {
   isbn10: string | null;
@@ -40,9 +40,17 @@ export type MetadataSearchQuery = {
 
 export type MetadataMatch = Partial<BookMetadata> & {
   sourceId?: string | null;
+  coverUrl?: string | null;
   identifiers?: Partial<BookIdentifiers>;
+  scoreTitle?: string | null;
 };
 
+export type MetadataField = "title" | "author" | "description" | "isbn" | "language" | "publisher" | "publishedDate" | "coverPath";
+
+export type MetadataStatus = "enriched" | "partial" | "none";
+
+export type MetadataProvenance = Partial<Record<string, MetadataProviderId | "existing" | string>>;
+
 export type MetadataProviderConfig = {
   provider: MetadataProviderId;
   enabled: boolean;
diff --git a/apps/api/src/metadata/normalize-published-date.test.ts b/apps/api/src/metadata/normalize-published-date.test.ts
new file mode 100644
index 0000000..5bd62f0
--- /dev/null
+++ b/apps/api/src/metadata/normalize-published-date.test.ts
@@ -0,0 +1,23 @@
+import { describe, expect, it } from "vitest";
+import { normalizePublishedDate } from "./use-cases/normalize-published-date.js";
+
+describe("normalizePublishedDate", () => {
+  it("rejects sentinel and absurd dates seen in real metadata providers", () => {
+    expect(normalizePublishedDate("0101-01-01T00:00:00+00:00")).toBeNull();
+    expect(normalizePublishedDate("0001-01-01")).toBeNull();
+    expect(normalizePublishedDate("1970-01-01")).toBeNull();
+    expect(normalizePublishedDate("0000")).toBeNull();
+  });
+
+  it("keeps only credible supported date formats", () => {
+    expect(normalizePublishedDate("2007")).toBe("2007");
+    expect(normalizePublishedDate("2007-07")).toBe("2007-07");
+    expect(normalizePublishedDate("2007-07-21")).toBe("2007-07-21");
+    expect(normalizePublishedDate("2007-07-21T00:00:00+00:00")).toBe("2007-07-21");
+  });
+
+  it("rejects years outside the supported publication range", () => {
+    expect(normalizePublishedDate("1499")).toBeNull();
+    expect(normalizePublishedDate("2028")).toBeNull();
+  });
+});
diff --git a/apps/api/src/metadata/use-cases/normalize-published-date.ts b/apps/api/src/metadata/use-cases/normalize-published-date.ts
new file mode 100644
index 0000000..9c7fe72
--- /dev/null
+++ b/apps/api/src/metadata/use-cases/normalize-published-date.ts
@@ -0,0 +1,44 @@
+const minimumYear = 1500;
+const maximumYear = 2027;
+const rejectedExactDates = new Set(["0001-01-01", "0101-01-01", "1970-01-01"]);
+
+export function normalizePublishedDate(value: string | null | undefined): string | null {
+  const text = value?.trim();
+  if (!text) return null;
+
+  const isoDate = text.match(/^(\d{4})-(\d{2})-(\d{2})(?:[T\s].*)?$/);
+  if (isoDate) {
+    const [, year, month, day] = isoDate;
+    const date = `${year}-${month}-${day}`;
+    if (rejectedExactDates.has(date)) return null;
+    return validDate(Number(year), Number(month), Number(day)) ? date : null;
+  }
+
+  const yearMonth = text.match(/^(\d{4})-(\d{2})$/);
+  if (yearMonth) {
+    const [, year, month] = yearMonth;
+    return validYear(Number(year)) && validMonth(Number(month)) ? `${year}-${month}` : null;
+  }
+
+  const yearOnly = text.match(/^(\d{4})$/);
+  if (yearOnly) {
+    const year = Number(yearOnly[1]);
+    return validYear(year) ? yearOnly[1] : null;
+  }
+
+  return null;
+}
+
+function validDate(year: number, month: number, day: number): boolean {
+  if (!validYear(year) || !validMonth(month) || day < 1 || day > 31) return false;
+  const date = new Date(Date.UTC(year, month - 1, day));
+  return date.getUTCFullYear() === year && date.getUTCMonth() === month - 1 && date.getUTCDate() === day;
+}
+
+function validYear(year: number): boolean {
+  return Number.isInteger(year) && year >= minimumYear && year <= maximumYear;
+}
+
+function validMonth(month: number): boolean {
+  return Number.isInteger(month) && month >= 1 && month <= 12;
+}
diff --git a/apps/api/src/metadata/use-cases/score-metadata-match.ts b/apps/api/src/metadata/use-cases/score-metadata-match.ts
index a81de00..84e16b3 100644
--- a/apps/api/src/metadata/use-cases/score-metadata-match.ts
+++ b/apps/api/src/metadata/use-cases/score-metadata-match.ts
@@ -3,56 +3,125 @@ import { MetadataMatch, MetadataSearchQuery } from "../metadata.types.js";
 export type ScoredMetadataMatch = {
   match: MetadataMatch;
   score: number;
+  titleScore: number;
+  authorScore: number | null;
+  dateScore: number | null;
+  isbnMatch: boolean;
 };
 
 export class ScoreMetadataMatch {
   score(query: MetadataSearchQuery, match: MetadataMatch): number {
-    let score = 0;
-    const titleScore = similarity(normalize(query.title), normalize(match.title ?? ""));
-    score += titleScore * 0.7;
-
-    if (query.author && match.author) {
-      score += similarity(normalize(query.author), normalize(match.author)) * 0.2;
-    } else if (!query.author) {
-      score += 0.08;
-    }
-
-    const queryYear = query.year ?? null;
-    const matchYear = match.publishedDate?.match(/\b(1[5-9]\d{2}|20\d{2})\b/)?.[1] ?? null;
-    if (queryYear && matchYear) score += queryYear === matchYear ? 0.1 : -0.1;
-
-    if (query.isbn && match.isbn && query.isbn.replace(/\D/g, "") === match.isbn.replace(/\D/g, "")) {
-      score += 0.25;
-    }
-    return Math.max(0, Math.min(1, score));
+    return this.details(query, match).score;
   }
 
-  best(query: MetadataSearchQuery, matches: MetadataMatch[], minimumScore = 0.55): ScoredMetadataMatch | null {
+  details(query: MetadataSearchQuery, match: MetadataMatch): ScoredMetadataMatch {
+    const titleScore = scoreTitle(query.title, match.scoreTitle ?? match.title ?? "");
+    const authorScore = query.author ? scoreAuthor(query.author, match.author) : null;
+    const dateScore = query.year ? scoreDate(query.year, match.publishedDate) : null;
+    const isbnMatch = exactIsbnMatch(query.isbn, match.isbn);
+    const maxPossible = 100 + (authorScore == null ? 0 : 30) + (dateScore == null ? 0 : 10);
+    const sum = titleScore + (authorScore ?? 0) + (dateScore ?? 0);
+    return {
+      match,
+      score: maxPossible ? Math.round((100 * sum) / maxPossible) : 0,
+      titleScore,
+      authorScore,
+      dateScore,
+      isbnMatch
+    };
+  }
+
+  best(query: MetadataSearchQuery, matches: MetadataMatch[], minimumScore = 0): ScoredMetadataMatch | null {
     const scored = matches
-      .map((match) => ({ match, score: this.score(query, match) }))
-      .sort((left, right) => right.score - left.score);
+      .map((match) => this.details(query, match))
+      .sort((left, right) => compareScoredMatches(query, left, right));
     const best = scored[0];
     return best && best.score >= minimumScore ? best : null;
   }
 }
 
-function normalize(value: string): string {
+function compareScoredMatches(query: MetadataSearchQuery, left: ScoredMetadataMatch, right: ScoredMetadataMatch): number {
+  if (left.isbnMatch !== right.isbnMatch) return left.isbnMatch ? -1 : 1;
+  const leftTitleAuthor = isHighConfidenceTitleAuthor(query, left);
+  const rightTitleAuthor = isHighConfidenceTitleAuthor(query, right);
+  if (leftTitleAuthor !== rightTitleAuthor) return leftTitleAuthor ? -1 : 1;
+  return right.score - left.score;
+}
+
+function isHighConfidenceTitleAuthor(query: MetadataSearchQuery, scored: ScoredMetadataMatch): boolean {
+  return scored.titleScore > 90 && (!query.author || (scored.authorScore ?? 0) >= 15);
+}
+
+function scoreTitle(left: string, right: string): number {
+  const normalizedLeft = normalizeTitle(left);
+  const normalizedRight = normalizeTitle(right);
+  if (!normalizedLeft || !normalizedRight) return 0;
+  if (normalizedLeft === normalizedRight) return 100;
+  if (normalizedLeft.includes(normalizedRight) || normalizedRight.includes(normalizedLeft)) return 95;
+  return Math.round(jaccard(tokens(normalizedLeft), tokens(normalizedRight)) * 100);
+}
+
+function scoreAuthor(localAuthor: string, sourceAuthor: string | null | undefined): number {
+  const local = authorSet(localAuthor);
+  if (!local.size) return 0;
+  const source = authorSet(sourceAuthor ?? "");
+  const present = [...local].filter((author) => source.has(author)).length;
+  return 30 * (present / local.size);
+}
+
+function scoreDate(localYear: string, sourceDate: string | null | undefined): number {
+  const left = Number(yearFrom(localYear));
+  const right = Number(yearFrom(sourceDate ?? ""));
+  if (!left || !right) return 0;
+  if (left === right) return 10;
+  return Math.abs(left - right) <= 1 ? 5 : 0;
+}
+
+function exactIsbnMatch(left: string | null | undefined, right: string | null | undefined): boolean {
+  const normalizedLeft = normalizeIsbn(left);
+  const normalizedRight = normalizeIsbn(right);
+  return Boolean(normalizedLeft && normalizedRight && normalizedLeft === normalizedRight);
+}
+
+function normalizeTitle(value: string): string {
+  return normalizeText(value.split(":")[0] ?? "").replace(/^(?:le|la|les|the|a|an|l)\s+/, "");
+}
+
+function normalizeText(value: string): string {
   return value
     .normalize("NFD")
     .replace(/[\u0300-\u036f]/g, "")
     .toLowerCase()
     .replace(/[^a-z0-9]+/g, " ")
-    .replace(/\b(le|la|les|the|a|an|de|du|des|et|and)\b/g, " ")
     .replace(/\s+/g, " ")
     .trim();
 }
 
-function similarity(left: string, right: string): number {
-  if (!left || !right) return 0;
-  if (left === right) return 1;
-  const leftTokens = new Set(left.split(" "));
-  const rightTokens = new Set(right.split(" "));
+function authorSet(value: string): Set<string> {
+  return new Set(
+    value
+      .split(/[,;&/]|\band\b|\bet\b/gi)
+      .map(normalizeText)
+      .filter(Boolean)
+      .sort()
+  );
+}
+
+function tokens(value: string): Set<string> {
+  return new Set(value.split(" ").filter(Boolean));
+}
+
+function jaccard(leftTokens: Set<string>, rightTokens: Set<string>): number {
   const intersection = [...leftTokens].filter((token) => rightTokens.has(token)).length;
   const union = new Set([...leftTokens, ...rightTokens]).size;
   return union ? intersection / union : 0;
 }
+
+function yearFrom(value: string): string | null {
+  return value.match(/\b(1[5-9]\d{2}|20\d{2})\b/)?.[1] ?? null;
+}
+
+function normalizeIsbn(value: string | null | undefined): string | null {
+  const normalized = value?.replace(/[^0-9X]/gi, "").toUpperCase() ?? "";
+  return normalized || null;
+}
diff --git a/apps/api/src/scanner/metadata.test.ts b/apps/api/src/scanner/metadata.test.ts
index 18e0b9a..c7ca2b4 100644
--- a/apps/api/src/scanner/metadata.test.ts
+++ b/apps/api/src/scanner/metadata.test.ts
@@ -2,10 +2,15 @@ import { mkdtempSync, writeFileSync } from "node:fs";
 import { tmpdir } from "node:os";
 import { join } from "node:path";
 import AdmZip from "adm-zip";
-import { describe, expect, it } from "vitest";
+import { describe, expect, it, vi } from "vitest";
 import { listCbzImageEntries } from "../common/cbz.js";
 import { extractMetadata } from "./metadata.js";
 
+vi.mock("../common/cbr.js", () => ({
+  listCbrImageEntries: async () => [{ entryName: "001.jpg", name: "001.jpg" }],
+  readCbrPage: async () => ({ entryName: "001.jpg", data: Buffer.from([0xff, 0xd8, 0xff, 0xd9]) })
+}));
+
 describe("pdf metadata extraction", () => {
   it("falls back to file name and reads simple PDF info fields", async () => {
     const dir = mkdtempSync(join(tmpdir(), "readabook-"));
@@ -36,3 +41,16 @@ describe("cbz metadata extraction", () => {
     expect(pages.map((page) => page.name)).toEqual(["001.jpg", "002.jpg"]);
   });
 });
+
+describe("cbr metadata extraction", () => {
+  it("uses the file name as title and first extracted image as cover", async () => {
+    const dir = mkdtempSync(join(tmpdir(), "readabook-"));
+    const file = join(dir, "Comic Two.cbr");
+    writeFileSync(file, "rar");
+
+    const metadata = await extractMetadata(file, dir);
+
+    expect(metadata.title).toBe("Comic Two");
+    expect(metadata.coverPath).toMatch(/covers\/[a-f0-9]+\.jpg$/);
+  });
+});
diff --git a/apps/api/src/scanner/metadata.ts b/apps/api/src/scanner/metadata.ts
index ea232f1..b5bd46a 100644
--- a/apps/api/src/scanner/metadata.ts
+++ b/apps/api/src/scanner/metadata.ts
@@ -3,8 +3,9 @@ import { mkdirSync, readFileSync, writeFileSync } from "node:fs";
 import { basename, dirname, extname, join } from "node:path";
 import AdmZip from "adm-zip";
 import { XMLParser } from "fast-xml-parser";
-import { listCbrImageEntries } from "../common/cbr.js";
+import { listCbrImageEntries, readCbrPage } from "../common/cbr.js";
 import { listCbzImageEntries } from "../common/cbz.js";
+import { normalizePublishedDate } from "../metadata/use-cases/normalize-published-date.js";
 
 export type BookMetadata = {
   title: string;
@@ -64,7 +65,7 @@ function extractEpubMetadata(filePath: string, storageDir: string): BookMetadata
     isbn,
     language: firstText(metadata["dc:language"]),
     publisher: firstText(metadata["dc:publisher"]),
-    publishedDate: firstText(metadata["dc:date"]),
+    publishedDate: normalizePublishedDate(firstText(metadata["dc:date"])),
     coverPath
   };
 }
@@ -97,10 +98,12 @@ function extractCbzMetadata(filePath: string, storageDir: string): BookMetadata
 }
 
 async function extractCbrMetadata(filePath: string, storageDir: string): Promise<BookMetadata> {
-  await listCbrImageEntries(filePath);
+  const firstPage = (await listCbrImageEntries(filePath))[0];
+  const page = await readCbrPage(filePath, 1, storageDir);
+  const coverPath = writeCoverData(page.data, firstPage.entryName, filePath, storageDir);
   return {
     ...fallbackMetadata(filePath),
-    coverPath: null
+    coverPath
   };
 }
 
@@ -165,6 +168,15 @@ function extractCover(zip: AdmZip, coverPathInZip: string, filePath: string, sto
   return target;
 }
 
+function writeCoverData(data: Buffer, entryName: string, filePath: string, storageDir: string): string {
+  const extension = extname(entryName) || ".jpg";
+  const hash = createHash("sha256").update(filePath).digest("hex").slice(0, 24);
+  const target = join(storageDir, "covers", `${hash}${extension}`);
+  mkdirSync(dirname(target), { recursive: true });
+  writeFileSync(target, data);
+  return target;
+}
+
 function matchPdfInfo(text: string, key: string): string | null {
   return text.match(new RegExp(`/${key}\\s*\\(([^)]{1,500})\\)`))?.[1] ?? null;
 }
diff --git a/apps/api/src/scanner/scanner.service.test.ts b/apps/api/src/scanner/scanner.service.test.ts
index 2238eee..b23b2b8 100644
--- a/apps/api/src/scanner/scanner.service.test.ts
+++ b/apps/api/src/scanner/scanner.service.test.ts
@@ -1,5 +1,24 @@
-import { describe, expect, it } from "vitest";
-import { scanDigest } from "./scanner.service.js";
+import { mkdtempSync, rmSync } from "node:fs";
+import { tmpdir } from "node:os";
+import { join } from "node:path";
+import { afterEach, describe, expect, it } from "vitest";
+import { DatabaseService } from "../database/database.service.js";
+import { books, libraries } from "../database/schema.js";
+import { JobsService } from "../jobs/jobs.service.js";
+import { enrichmentDigest, preserveExistingBookValues, scanDigest } from "./scanner.service.js";
+import { ScannerService } from "./scanner.service.js";
+
+const previousDatabasePath = process.env.DATABASE_PATH;
+const previousStorageDir = process.env.STORAGE_DIR;
+const tempDirs: string[] = [];
+
+afterEach(() => {
+  process.env.DATABASE_PATH = previousDatabasePath;
+  process.env.STORAGE_DIR = previousStorageDir;
+  for (const dir of tempDirs.splice(0)) {
+    rmSync(dir, { recursive: true, force: true });
+  }
+});
 
 describe("scan digest", () => {
   it("reports incomplete files without exposing huge traces", () => {
@@ -17,4 +36,217 @@ describe("scan digest", () => {
     expect(detail).toContain("broken.cbr");
     expect(detail.length).toBeLessThan(380);
   });
+
+  it("reports metadata enrichment jobs as enrichment, not scans", () => {
+    expect(enrichmentDigest(35, [])).toBe("Enriched 35 book(s)");
+  });
+
+  it("does not erase existing metadata or cover when a rescan has less information", () => {
+    const existing: typeof books.$inferSelect = {
+      id: 1,
+      libraryId: 1,
+      seriesId: null,
+      title: "Daredevil",
+      author: "Roy Thomas",
+      description: "Existing description",
+      isbn: "9782809476255",
+      isbn13: "9782809476255",
+      identifiersJson: null,
+      localMetadataJson: null,
+      language: "fre",
+      publisher: "Panini comics",
+      publishedDate: "2019",
+      volumeNumber: null,
+      volumeLabel: null,
+      format: "cbz",
+      filePath: "/library/Daredevil.cbz",
+      coverPath: "/storage/covers/daredevil.jpg",
+      metadataStatus: "enriched",
+      metadataProvenanceJson: JSON.stringify({ author: "bnf", coverPath: "openlibrary" }),
+      scanStatus: "succeeded",
+      enrichmentStatus: "succeeded",
+      fileSize: 12,
+      fileMtime: "2026-08-23T00:00:00.000Z",
+      createdAt: "2026-08-23T00:00:00.000Z",
+      updatedAt: "2026-08-23T00:00:00.000Z"
+    };
+
+    const next = preserveExistingBookValues(
+      {
+        title: "Daredevil",
+        author: null,
+        description: null,
+        isbn: null,
+        isbn13: null,
+        language: null,
+        publisher: null,
+        publishedDate: null,
+        coverPath: null,
+        metadataStatus: "none",
+        metadataProvenanceJson: JSON.stringify({ title: "local" }),
+        scanStatus: "succeeded" as const
+      },
+      existing
+    );
+
+    expect(next).toMatchObject({
+      author: "Roy Thomas",
+      description: "Existing description",
+      isbn: "9782809476255",
+      coverPath: "/storage/covers/daredevil.jpg",
+      metadataStatus: "enriched"
+    });
+    expect(JSON.parse(String(next.metadataProvenanceJson))).toMatchObject({
+      title: "local",
+      author: "bnf",
+      coverPath: "openlibrary"
+    });
+  });
+
+  it("does not replace an existing valid publication date with a sentinel date", () => {
+    const existing = {
+      id: 1,
+      libraryId: 1,
+      seriesId: null,
+      title: "Lord of the Mysteries",
+      author: null,
+      description: null,
+      isbn: null,
+      isbn13: null,
+      identifiersJson: null,
+      localMetadataJson: null,
+      language: null,
+      publisher: null,
+      publishedDate: "2018",
+      volumeNumber: null,
+      volumeLabel: null,
+      format: "epub",
+      filePath: "/library/Lord of the Mysteries.epub",
+      coverPath: null,
+      metadataStatus: "partial",
+      metadataProvenanceJson: JSON.stringify({ publishedDate: "existing" }),
+      scanStatus: "succeeded",
+      enrichmentStatus: "succeeded",
+      fileSize: 12,
+      fileMtime: "2026-08-23T00:00:00.000Z",
+      createdAt: "2026-08-23T00:00:00.000Z",
+      updatedAt: "2026-08-23T00:00:00.000Z"
+    } satisfies typeof books.$inferSelect;
+
+    const next = preserveExistingBookValues(
+      {
+        title: "Lord of the Mysteries",
+        publishedDate: "0101-01-01T00:00:00+00:00",
+        metadataStatus: "partial",
+        metadataProvenanceJson: JSON.stringify({ title: "local", publishedDate: "openlibrary" })
+      },
+      existing
+    );
+
+    expect(next.publishedDate).toBe("2018");
+  });
+
+  it.runIf(canLoadBetterSqlite())("updates the existing book when an insert races with books.file_path uniqueness", () => {
+    const database = createDatabase();
+    const now = database.now();
+    const library = database.db
+      .insert(libraries)
+      .values({ name: "Corpus", path: "/library", enabled: true, createdAt: now, updatedAt: now })
+      .returning()
+      .get();
+    database.db
+      .insert(books)
+      .values({
+        libraryId: library.id,
+        seriesId: null,
+        title: "Daredevil",
+        author: null,
+        description: null,
+        isbn: null,
+        isbn13: null,
+        identifiersJson: null,
+        localMetadataJson: null,
+        language: null,
+        publisher: null,
+        publishedDate: null,
+        volumeNumber: null,
+        volumeLabel: null,
+        format: "cbz",
+        filePath: "/library/Daredevil.cbz",
+        coverPath: null,
+        metadataStatus: "none",
+        metadataProvenanceJson: null,
+        scanStatus: "succeeded",
+        enrichmentStatus: "succeeded",
+        fileSize: 1,
+        fileMtime: now,
+        createdAt: now,
+        updatedAt: now
+      })
+      .run();
+    const scanner = new ScannerService(database, new JobsService(database), {} as never);
+    const values: Omit<typeof books.$inferInsert, "createdAt"> = {
+      libraryId: library.id,
+      seriesId: null,
+      title: "Daredevil",
+      author: "Roy Thomas",
+      description: "Updated metadata",
+      isbn: null,
+      isbn13: null,
+      identifiersJson: null,
+      localMetadataJson: null,
+      language: null,
+      publisher: null,
+      publishedDate: null,
+      volumeNumber: null,
+      volumeLabel: null,
+      format: "cbz",
+      filePath: "/library/Daredevil.cbz",
+      coverPath: "/storage/covers/daredevil.jpg",
+      metadataStatus: "enriched",
+      metadataProvenanceJson: JSON.stringify({ author: "bnf", coverPath: "local" }),
+      scanStatus: "succeeded",
+      enrichmentStatus: "succeeded",
+      fileSize: 2,
+      fileMtime: now,
+      updatedAt: now
+    };
+
+    (scanner as unknown as {
+      upsertBookByFilePath(values: Omit<typeof books.$inferInsert, "createdAt">, existing: undefined, createdAt: string): void;
+    }).upsertBookByFilePath(
+      values,
+      undefined,
+      now
+    );
+
+    const rows = database.db.select().from(books).all();
+    expect(rows).toHaveLength(1);
+    expect(rows[0]).toMatchObject({
+      author: "Roy Thomas",
+      coverPath: "/storage/covers/daredevil.jpg",
+      metadataStatus: "enriched",
+      fileSize: 2
+    });
+
+    database.onModuleDestroy();
+  });
 });
+
+function createDatabase(): DatabaseService {
+  const dir = mkdtempSync(join(tmpdir(), "readabook-scanner-service-"));
+  tempDirs.push(dir);
+  process.env.DATABASE_PATH = join(dir, "readabook.sqlite");
+  process.env.STORAGE_DIR = join(dir, "storage");
+  return new DatabaseService();
+}
+
+function canLoadBetterSqlite(): boolean {
+  try {
+    const database = createDatabase();
+    database.onModuleDestroy();
+    return true;
+  } catch {
+    return false;
+  }
+}
diff --git a/apps/api/src/scanner/scanner.service.ts b/apps/api/src/scanner/scanner.service.ts
index b526caa..33382b6 100644
--- a/apps/api/src/scanner/scanner.service.ts
+++ b/apps/api/src/scanner/scanner.service.ts
@@ -3,9 +3,11 @@ import { existsSync, readdirSync, statSync } from "node:fs";
 import { basename, extname, join } from "node:path";
 import { eq, inArray } from "drizzle-orm";
 import { DatabaseService } from "../database/database.service.js";
-import { automationSettings, books, libraries } from "../database/schema.js";
+import { automationSettings, books, libraries, series } from "../database/schema.js";
 import { JobsService } from "../jobs/jobs.service.js";
 import { MetadataService } from "../metadata/metadata.service.js";
+import { extractSeriesVolume } from "../metadata/use-cases/extract-series-volume.js";
+import { normalizePublishedDate } from "../metadata/use-cases/normalize-published-date.js";
 import { extractMetadata } from "./metadata.js";
 
 @Injectable()
@@ -86,23 +88,40 @@ export class ScannerService {
     this.jobs.markRunning(jobId, "Enriching existing books");
     const rows = this.database.db.select({ id: books.id }).from(books).all();
     let count = 0;
+    const failures: ScanFailure[] = [];
     for (const row of rows) {
-      await this.metadata.enrichBook(row.id);
-      count += 1;
+      this.markBookEnrichmentStatus(row.id, "running");
+      try {
+        await this.metadata.enrichBook(row.id);
+        this.markBookEnrichmentStatus(row.id, "succeeded");
+        count += 1;
+      } catch (error) {
+        this.markBookEnrichmentStatus(row.id, "failed");
+        failures.push({ filePath: `book #${row.id}`, error: errorMessage(error) });
+      }
     }
-    this.jobs.markSucceeded(jobId, `Enriched ${count} book(s)`);
+    this.jobs.markSucceeded(jobId, enrichmentDigest(count, failures));
   }
 
   private async ingestFile(libraryId: number, filePath: string): Promise<void> {
     const stats = statSync(filePath);
+    const existing = this.database.db.select().from(books).where(eq(books.filePath, filePath)).get();
+    if (existing) {
+      this.database.db
+        .update(books)
+        .set({ scanStatus: "running", enrichmentStatus: "running", updatedAt: this.database.now() })
+        .where(eq(books.id, existing.id))
+        .run();
+    }
     const localMetadata = await extractMetadata(filePath, this.database.config.storageDir);
     const now = this.database.now();
     const format = bookFormatFromPath(filePath);
-    const existing = this.database.db.select({ id: books.id }).from(books).where(eq(books.filePath, filePath)).get();
     const shouldRemoteEnrich = !existing ? this.shouldAutoEnrichNewBooks() : true;
     const metadata = await this.metadata.enrichMetadata(localMetadata, filePath, { remote: shouldRemoteEnrich });
+    const seriesInfo = this.resolveSeries(metadata.title, filePath);
     const values = {
       libraryId,
+      seriesId: seriesInfo.seriesId,
       title: metadata.title,
       author: metadata.author,
       description: metadata.description,
@@ -113,25 +132,31 @@ export class ScannerService {
       language: metadata.language,
       publisher: metadata.publisher,
       publishedDate: metadata.publishedDate,
+      volumeNumber: seriesInfo.volumeNumber,
+      volumeLabel: seriesInfo.volumeLabel,
       format,
       filePath,
       coverPath: metadata.coverPath,
+      metadataStatus: metadata.metadataStatus,
+      metadataProvenanceJson: metadata.metadataProvenanceJson,
+      scanStatus: "succeeded" as const,
+      enrichmentStatus: shouldRemoteEnrich ? ("succeeded" as const) : ("idle" as const),
       fileSize: stats.size,
       fileMtime: stats.mtime.toISOString(),
       updatedAt: now
     };
 
-    existing
-      ? this.database.db.update(books).set(values).where(eq(books.id, existing.id)).returning().get()
-      : this.database.db.insert(books).values({ ...values, createdAt: now }).returning().get();
+    this.upsertBookByFilePath(values, existing, now);
   }
 
   private ingestIncompleteFile(libraryId: number, filePath: string): void {
     const stats = statSync(filePath);
     const now = this.database.now();
-    const existing = this.database.db.select({ id: books.id }).from(books).where(eq(books.filePath, filePath)).get();
+    const existing = this.database.db.select().from(books).where(eq(books.filePath, filePath)).get();
+    const seriesInfo = this.resolveSeries(basename(filePath, extname(filePath)), filePath);
     const values = {
       libraryId,
+      seriesId: seriesInfo.seriesId,
       title: basename(filePath, extname(filePath)),
       author: null,
       description: null,
@@ -154,17 +179,41 @@ export class ScannerService {
       language: null,
       publisher: null,
       publishedDate: null,
+      volumeNumber: seriesInfo.volumeNumber,
+      volumeLabel: seriesInfo.volumeLabel,
       format: bookFormatFromPath(filePath),
       filePath,
       coverPath: null,
+      metadataStatus: "none" as const,
+      metadataProvenanceJson: JSON.stringify({ title: "local" }),
+      scanStatus: "failed" as const,
+      enrichmentStatus: "failed" as const,
       fileSize: stats.size,
       fileMtime: stats.mtime.toISOString(),
       updatedAt: now
     };
 
-    existing
-      ? this.database.db.update(books).set(values).where(eq(books.id, existing.id)).returning().get()
-      : this.database.db.insert(books).values({ ...values, createdAt: now }).returning().get();
+    this.upsertBookByFilePath(values, existing, now);
+  }
+
+  private upsertBookByFilePath(
+    values: Omit<typeof books.$inferInsert, "createdAt">,
+    existing: typeof books.$inferSelect | undefined,
+    createdAt: string
+  ): void {
+    if (existing) {
+      this.database.db.update(books).set(preserveExistingBookValues(values, existing)).where(eq(books.id, existing.id)).run();
+      return;
+    }
+    try {
+      this.database.db.insert(books).values({ ...values, createdAt }).run();
+      return;
+    } catch (error) {
+      if (!isUniqueFilePathError(error)) throw error;
+      const current = this.database.db.select().from(books).where(eq(books.filePath, values.filePath)).get();
+      if (!current) throw error;
+      this.database.db.update(books).set(preserveExistingBookValues(values, current)).where(eq(books.id, current.id)).run();
+    }
   }
 
   private removeMissingBooks(libraryId: number, seen: Set<string>): number {
@@ -184,6 +233,32 @@ export class ScannerService {
         .get()?.autoEnrichNewBooks
     );
   }
+
+  private markBookEnrichmentStatus(id: number, enrichmentStatus: "running" | "succeeded" | "failed"): void {
+    this.database.db.update(books).set({ enrichmentStatus, updatedAt: this.database.now() }).where(eq(books.id, id)).run();
+  }
+
+  private resolveSeries(title: string, filePath: string): { seriesId: number; volumeNumber: number | null; volumeLabel: string | null } {
+    const parsed = extractSeriesVolume(title, filePath);
+    const now = this.database.now();
+    const row = this.database.db
+      .insert(series)
+      .values({
+        title: parsed.seriesTitle,
+        normalizedTitle: parsed.normalizedSeriesTitle,
+        description: null,
+        publisher: null,
+        createdAt: now,
+        updatedAt: now
+      })
+      .onConflictDoUpdate({
+        target: series.normalizedTitle,
+        set: { title: parsed.seriesTitle, updatedAt: now }
+      })
+      .returning({ id: series.id })
+      .get();
+    return { seriesId: row.id, volumeNumber: parsed.volumeNumber, volumeLabel: parsed.volumeLabel };
+  }
 }
 
 type ScanFailure = {
@@ -202,6 +277,17 @@ export function scanDigest(scanned: number, removed: number, failures: ScanFailu
   return `${base}, ${failures.length} incomplete file(s): ${examples}${extra}`;
 }
 
+export function enrichmentDigest(enriched: number, failures: ScanFailure[]): string {
+  const base = `Enriched ${enriched} book(s)`;
+  if (!failures.length) return base;
+  const examples = failures
+    .slice(0, 3)
+    .map((failure) => `${basename(failure.filePath)}: ${truncate(failure.error)}`)
+    .join("; ");
+  const extra = failures.length > 3 ? `; ${failures.length - 3} more` : "";
+  return `${base}, ${failures.length} incomplete book(s): ${examples}${extra}`;
+}
+
 function errorMessage(error: unknown): string {
   if (error instanceof Error && error.message) return truncate(error.message);
   return truncate(String(error));
@@ -233,3 +319,49 @@ function bookFormatFromPath(filePath: string): "epub" | "pdf" | "cbz" | "cbr" {
   if (extension === ".cbr") return "cbr";
   return "pdf";
 }
+
+export function preserveExistingBookValues<T extends Partial<typeof books.$inferInsert>>(values: T, existing: typeof books.$inferSelect): T {
+  const next = { ...values };
+  if (next.scanStatus === "failed" && existing.title) {
+    next.title = existing.title as never;
+  }
+  for (const field of ["author", "description", "isbn", "isbn13", "language", "publisher", "publishedDate", "coverPath", "seriesId", "volumeNumber", "volumeLabel"] as const) {
+    if (field === "publishedDate") {
+      next.publishedDate = (normalizePublishedDate(next.publishedDate) ?? normalizePublishedDate(existing.publishedDate)) as never;
+      continue;
+    }
+    if (next[field] == null && existing[field] != null) {
+      next[field] = existing[field] as never;
+    }
+  }
+  next.metadataStatus = computeMetadataStatus(next, existing.metadataStatus) as never;
+  next.metadataProvenanceJson = mergeProvenanceJson(String(next.metadataProvenanceJson ?? "{}"), existing.metadataProvenanceJson) as never;
+  return next;
+}
+
+function computeMetadataStatus(values: Partial<typeof books.$inferInsert>, existingStatus: string): "enriched" | "partial" | "none" {
+  const hasCover = Boolean(values.coverPath);
+  const filled = [values.author, values.description, values.isbn, values.language, values.publisher, values.publishedDate].filter(Boolean).length;
+  const computed = hasCover && filled >= 2 ? "enriched" : hasCover || filled > 0 ? "partial" : "none";
+  const rank = { none: 0, partial: 1, enriched: 2 } as const;
+  const safeExisting = existingStatus === "enriched" || existingStatus === "partial" || existingStatus === "none" ? existingStatus : "none";
+  return rank[computed] >= rank[safeExisting] ? computed : safeExisting;
+}
+
+function mergeProvenanceJson(nextJson: string, existingJson: string | null): string {
+  return JSON.stringify({ ...parseJsonObject(existingJson), ...parseJsonObject(nextJson) });
+}
+
+function parseJsonObject(value: string | null): Record<string, unknown> {
+  if (!value) return {};
+  try {
+    const parsed = JSON.parse(value) as unknown;
+    return parsed && typeof parsed === "object" && !Array.isArray(parsed) ? (parsed as Record<string, unknown>) : {};
+  } catch {
+    return {};
+  }
+}
+
+function isUniqueFilePathError(error: unknown): boolean {
+  return error instanceof Error && error.message.includes("UNIQUE constraint failed: books.file_path");
+}
diff --git a/apps/web/src/pages/AdminAutomationPage.tsx b/apps/web/src/pages/AdminAutomationPage.tsx
index a570abf..840feef 100644
--- a/apps/web/src/pages/AdminAutomationPage.tsx
+++ b/apps/web/src/pages/AdminAutomationPage.tsx
@@ -4,16 +4,21 @@ import type {
   AutomationFrequency,
   AutomationScheduleDto,
   AutomationSettingsDto,
-  MetadataProviderId,
   MetadataSourcesConfigDto
 } from "@readabook/shared";
 import { api, getApiFallback } from "../api/client";
 import { ErrorRibbon, LoadingState, Panel } from "../components/ui";
 import {
+  type AdminMetadataProviderId,
+  type AdminMetadataSourcesConfig,
+  defaultMetadataSources,
   metadataSourcesPayload,
   moveSource,
   normalizeMetadataSources,
   providerLabels,
+  providerUiMessage,
+  providerUiState,
+  providerUiStateLabel,
   scheduleDays,
   scheduleSummary
 } from "./adminAutomation";
@@ -28,14 +33,9 @@ type ApiState<T> = {
   success?: string;
 };
 
-const defaultMetadataConfig: MetadataSourcesConfigDto = {
+const defaultMetadataConfig: AdminMetadataSourcesConfig = {
   isbnPriorityEnabled: true,
-  sources: [
-    { provider: "local", enabled: true, priority: 0, hasApiKey: false },
-    { provider: "openlibrary", enabled: false, priority: 1, hasApiKey: false },
-    { provider: "googlebooks", enabled: false, priority: 2, hasApiKey: false },
-    { provider: "bnf", enabled: false, priority: 3, hasApiKey: false }
-  ]
+  sources: defaultMetadataSources
 };
 
 const defaultAutomationSettings: AutomationSettingsDto = {
@@ -47,7 +47,7 @@ const defaultAutomationSettings: AutomationSettingsDto = {
 
 export function AdminAutomationPage() {
   const [tab, setTab] = useState<AdminAutomationTab>("sources");
-  const [metadataState, setMetadataState] = useState<ApiState<MetadataSourcesConfigDto>>({
+  const [metadataState, setMetadataState] = useState<ApiState<AdminMetadataSourcesConfig>>({
     initial: null,
     draft: null,
     loading: true,
@@ -59,7 +59,7 @@ export function AdminAutomationPage() {
     loading: true,
     saving: false
   });
-  const [apiKeys, setApiKeys] = useState<Partial<Record<MetadataProviderId, string>>>({});
+  const [apiKeys, setApiKeys] = useState<Partial<Record<AdminMetadataProviderId, string>>>({});
 
   const metadataDirty = useMemo(
     () => Boolean(metadataState.initial && metadataState.draft && JSON.stringify(metadataState.initial) !== JSON.stringify(metadataState.draft)),
@@ -78,7 +78,7 @@ export function AdminAutomationPage() {
       setMetadataState({ initial: next, draft: next, loading: false, saving: false });
       setApiKeys({});
     } catch (error) {
-      const fallback = getApiFallback<MetadataSourcesConfigDto>(error);
+      const fallback = getApiFallback<MetadataSourcesConfigDto | AdminMetadataSourcesConfig>(error);
       const next = normalizeMetadataSources(fallback ?? defaultMetadataConfig);
       setMetadataState({
         initial: next,
@@ -226,11 +226,11 @@ function MetadataSourcesPanel({
   onSubmit,
   onRefresh
 }: {
-  state: ApiState<MetadataSourcesConfigDto>;
+  state: ApiState<AdminMetadataSourcesConfig>;
   dirty: boolean;
-  apiKeys: Partial<Record<MetadataProviderId, string>>;
-  setApiKeys: (next: Partial<Record<MetadataProviderId, string>>) => void;
-  onChange: (draft: MetadataSourcesConfigDto) => void;
+  apiKeys: Partial<Record<AdminMetadataProviderId, string>>;
+  setApiKeys: (next: Partial<Record<AdminMetadataProviderId, string>>) => void;
+  onChange: (draft: AdminMetadataSourcesConfig) => void;
   onSubmit: (event: FormEvent) => void;
   onRefresh: () => Promise<void>;
 }) {
@@ -297,14 +297,25 @@ function MetadataSourcesPanel({
                   <small>{source.enabled ? "active" : "inactive"}</small>
                 </span>
               </label>
-              <label>
-                Cle API
-                <input
-                  value={apiKeys[source.provider] ?? ""}
-                  onChange={(event) => setApiKeys({ ...apiKeys, [source.provider]: event.target.value })}
-                  placeholder={source.hasApiKey ? "cle conservee" : "optionnelle"}
-                />
-              </label>
+              <div className="provider-config">
+                <div className="provider-state-line">
+                  <span className={`status-pill provider-state-${providerUiState(source)}`}>{providerUiStateLabel(source)}</span>
+                  <small>{providerUiMessage(source)}</small>
+                </div>
+                {source.provider === "comicvine" && (
+                  <p className="provider-warning">
+                    Usage non commercial uniquement. Verifier la compatibilite avec l'usage de ReadaBook.
+                  </p>
+                )}
+                <label>
+                  Cle API
+                  <input
+                    value={apiKeys[source.provider] ?? ""}
+                    onChange={(event) => setApiKeys({ ...apiKeys, [source.provider]: event.target.value })}
+                    placeholder={source.hasApiKey ? "cle conservee" : source.requiresCredentials ? "requise" : "optionnelle"}
+                  />
+                </label>
+              </div>
               <div className="provider-actions">
                 <button
                   className="ghost-button icon-button"
diff --git a/apps/web/src/pages/AdminPage.tsx b/apps/web/src/pages/AdminPage.tsx
index d3573dc..6f72640 100644
--- a/apps/web/src/pages/AdminPage.tsx
+++ b/apps/web/src/pages/AdminPage.tsx
@@ -2,8 +2,13 @@ import { FormEvent, useEffect, useState } from "react";
 import { Play, Plus, Trash2 } from "lucide-react";
 import type { JobDto, LibraryDto, UserDto } from "@readabook/shared";
 import { api, getApiFallback } from "../api/client";
+import { jobDigestSummary } from "../book/metadata";
 import { EmptyState, ErrorRibbon, LoadingState, Panel } from "../components/ui";
 
+function formatJobTime(value: string) {
+  return new Intl.DateTimeFormat(undefined, { hour: "2-digit", minute: "2-digit" }).format(new Date(value));
+}
+
 export function AdminPage() {
   const [libraries, setLibraries] = useState<LibraryDto[]>([]);
   const [jobs, setJobs] = useState<JobDto[]>([]);
@@ -161,7 +166,13 @@ export function AdminPage() {
           <div className="job-list">
             {jobs.map((job) => (
               <div key={job.id}>
-                <strong>{job.type}</strong>
+                <div className="job-copy">
+                  <div>
+                    <strong>{job.type}</strong>
+                    <small>{jobDigestSummary(job)}</small>
+                  </div>
+                  <time dateTime={job.updatedAt}>{formatJobTime(job.updatedAt)}</time>
+                </div>
                 <span>{job.status}</span>
               </div>
             ))}
diff --git a/apps/web/src/pages/adminAutomation.test.ts b/apps/web/src/pages/adminAutomation.test.ts
index 46d961e..e8ae5c6 100644
--- a/apps/web/src/pages/adminAutomation.test.ts
+++ b/apps/web/src/pages/adminAutomation.test.ts
@@ -1,14 +1,24 @@
 import { describe, expect, it } from "vitest";
-import type { MetadataSourcesConfigDto } from "@readabook/shared";
-import { metadataSourcesPayload, moveSource, normalizeMetadataSources, scheduleSummary } from "./adminAutomation";
+import type { AdminMetadataSourcesConfig } from "./adminAutomation";
+import {
+  metadataSourcesPayload,
+  moveSource,
+  normalizeMetadataSources,
+  providerLabels,
+  providerUiMessage,
+  providerUiStateLabel,
+  scheduleSummary
+} from "./adminAutomation";
 
-const config: MetadataSourcesConfigDto = {
+const config: AdminMetadataSourcesConfig = {
   isbnPriorityEnabled: true,
   sources: [
     { provider: "googlebooks", enabled: false, priority: 2, hasApiKey: true },
     { provider: "local", enabled: false, priority: 99, hasApiKey: false },
     { provider: "openlibrary", enabled: true, priority: 1, hasApiKey: false },
-    { provider: "bnf", enabled: false, priority: 3, hasApiKey: false }
+    { provider: "bnf", enabled: false, priority: 3, hasApiKey: false },
+    { provider: "comicvine", enabled: true, priority: 4, hasApiKey: false, requiresCredentials: true },
+    { provider: "mangadex", enabled: false, priority: 5, hasApiKey: false }
   ]
 };
 
@@ -27,17 +37,44 @@ describe("admin automation helpers", () => {
       sources: [
         { provider: "openlibrary", enabled: true, priority: 1 },
         { provider: "googlebooks", enabled: false, priority: 2 },
-        { provider: "bnf", enabled: false, priority: 3 }
+        { provider: "bnf", enabled: false, priority: 3 },
+        { provider: "comicvine", enabled: true, priority: 4 },
+        { provider: "mangadex", enabled: false, priority: 5 }
       ]
     });
   });
 
   it("moves only external providers", () => {
     const moved = moveSource(normalizeMetadataSources(config).sources, "bnf", -1);
-    expect(moved.map((source) => source.provider)).toEqual(["local", "openlibrary", "bnf", "googlebooks"]);
+    expect(moved.map((source) => source.provider)).toEqual(["local", "openlibrary", "bnf", "googlebooks", "comicvine", "mangadex"]);
   });
 
   it("summarizes weekly schedules", () => {
     expect(scheduleSummary({ frequency: "weekly", time: "04:30", dayOfWeek: 1 }, "Scan")).toBe("Scan chaque lundi a 04:30.");
   });
+
+  it("adds Comic Vine and MangaDex when the backend omits them", () => {
+    const normalized = normalizeMetadataSources({
+      isbnPriorityEnabled: true,
+      sources: [{ provider: "local", enabled: true, priority: 0, hasApiKey: false }]
+    });
+    expect(normalized.sources.map((source) => source.provider)).toContain("comicvine");
+    expect(normalized.sources.map((source) => source.provider)).toContain("mangadex");
+    expect(providerLabels.comicvine).toBe("Comic Vine");
+    expect(providerLabels.mangadex).toBe("MangaDex");
+  });
+
+  it("labels provider configuration, rate limit and error states", () => {
+    expect(providerUiStateLabel({ provider: "comicvine", enabled: true, priority: 1, hasApiKey: false, requiresCredentials: true })).toBe(
+      "A configurer"
+    );
+    expect(providerUiMessage({ provider: "comicvine", enabled: true, priority: 1, hasApiKey: false, requiresCredentials: true })).toBe(
+      "Source activee, configuration incomplete."
+    );
+    expect(providerUiStateLabel({ provider: "mangadex", enabled: true, priority: 2, hasApiKey: false, rateLimited: true })).toBe("Limite");
+    expect(providerUiMessage({ provider: "mangadex", enabled: true, priority: 2, hasApiKey: false, status: "quota_exceeded" })).toBe(
+      "Quota ou limite temporaire atteint. ReadaBook reessaiera plus tard."
+    );
+    expect(providerUiStateLabel({ provider: "mangadex", enabled: true, priority: 2, hasApiKey: false, lastError: "500 stack" })).toBe("Erreur");
+  });
 });
diff --git a/apps/web/src/pages/adminAutomation.ts b/apps/web/src/pages/adminAutomation.ts
index ba31007..5931e94 100644
--- a/apps/web/src/pages/adminAutomation.ts
+++ b/apps/web/src/pages/adminAutomation.ts
@@ -6,18 +6,57 @@ import type {
   UpdateMetadataSourcesConfigDto
 } from "@readabook/shared";
 
-export const providerLabels: Record<MetadataProviderId, string> = {
+export type AdminMetadataProviderId = MetadataProviderId | "comicvine" | "mangadex";
+
+export type AdminMetadataSourceConfig = Omit<MetadataSourceConfigDto, "provider"> & {
+  provider: AdminMetadataProviderId;
+  requiresCredentials?: boolean;
+  status?: string | null;
+  state?: string | null;
+  health?: string | null;
+  message?: string | null;
+  lastError?: string | null;
+  rateLimited?: boolean;
+  quotaLimited?: boolean;
+};
+
+export type AdminMetadataSourcesConfig = Omit<MetadataSourcesConfigDto, "sources"> & {
+  sources: AdminMetadataSourceConfig[];
+};
+
+export type ProviderUiState = "configured" | "missing-config" | "limited" | "error";
+
+export const providerLabels: Record<AdminMetadataProviderId, string> = {
   local: "Fichier local",
   openlibrary: "OpenLibrary",
   googlebooks: "Google Books",
-  bnf: "BnF"
+  bnf: "BnF",
+  comicvine: "Comic Vine",
+  mangadex: "MangaDex"
 };
 
+export const defaultMetadataSources: AdminMetadataSourceConfig[] = [
+  { provider: "local", enabled: true, priority: 0, hasApiKey: false },
+  { provider: "openlibrary", enabled: false, priority: 1, hasApiKey: false },
+  { provider: "googlebooks", enabled: false, priority: 2, hasApiKey: false },
+  { provider: "bnf", enabled: false, priority: 3, hasApiKey: false },
+  { provider: "comicvine", enabled: false, priority: 4, hasApiKey: false, requiresCredentials: true },
+  { provider: "mangadex", enabled: false, priority: 5, hasApiKey: false }
+];
+
 const weekdays = ["dimanche", "lundi", "mardi", "mercredi", "jeudi", "vendredi", "samedi"];
 
-export function normalizeMetadataSources(config: MetadataSourcesConfigDto): MetadataSourcesConfigDto {
-  const sorted = [...config.sources].sort((left, right) => left.priority - right.priority);
-  const local = sorted.find((source) => source.provider === "local") ?? { provider: "local", enabled: true, priority: 0, hasApiKey: false };
+export function normalizeMetadataSources(config: MetadataSourcesConfigDto | AdminMetadataSourcesConfig): AdminMetadataSourcesConfig {
+  const received = config.sources as AdminMetadataSourceConfig[];
+  const merged = defaultMetadataSources.map((source) => ({
+    ...source,
+    ...received.find((item) => item.provider === source.provider)
+  }));
+  received.forEach((source) => {
+    if (!merged.some((item) => item.provider === source.provider)) merged.push(source);
+  });
+  const sorted = merged.sort((left, right) => left.priority - right.priority);
+  const local = sorted.find((source) => source.provider === "local") ?? defaultMetadataSources[0];
   const external = sorted.filter((source) => source.provider !== "local");
   return {
     isbnPriorityEnabled: config.isbnPriorityEnabled,
@@ -28,8 +67,8 @@ export function normalizeMetadataSources(config: MetadataSourcesConfigDto): Meta
   };
 }
 
-export function metadataSourcesPayload(config: MetadataSourcesConfigDto): UpdateMetadataSourcesConfigDto {
-  const sources: NonNullable<UpdateMetadataSourcesConfigDto["sources"]> = [];
+export function metadataSourcesPayload(config: AdminMetadataSourcesConfig): UpdateMetadataSourcesConfigDto {
+  const sources: Array<{ provider: AdminMetadataProviderId; enabled: boolean; priority: number; apiKey?: string }> = [];
   config.sources.forEach((source) => {
     if (source.provider === "local") return;
     sources.push({
@@ -41,10 +80,10 @@ export function metadataSourcesPayload(config: MetadataSourcesConfigDto): Update
   return {
     isbnPriorityEnabled: config.isbnPriorityEnabled,
     sources
-  };
+  } as UpdateMetadataSourcesConfigDto;
 }
 
-export function moveSource(sources: MetadataSourceConfigDto[], provider: MetadataProviderId, direction: -1 | 1): MetadataSourceConfigDto[] {
+export function moveSource(sources: AdminMetadataSourceConfig[], provider: AdminMetadataProviderId, direction: -1 | 1): AdminMetadataSourceConfig[] {
   const external = sources.filter((source) => source.provider !== "local");
   const index = external.findIndex((source) => source.provider === provider);
   const nextIndex = index + direction;
@@ -56,6 +95,34 @@ export function moveSource(sources: MetadataSourceConfigDto[], provider: Metadat
   return [local, ...nextExternal].map((source, priority) => ({ ...source, priority: source.provider === "local" ? 0 : priority }));
 }
 
+function sourceStatusText(source: AdminMetadataSourceConfig): string {
+  return [source.status, source.state, source.health].filter(Boolean).join(" ").toLowerCase();
+}
+
+export function providerUiState(source: AdminMetadataSourceConfig): ProviderUiState {
+  const status = sourceStatusText(source);
+  if (source.rateLimited || source.quotaLimited || status.includes("limit") || status.includes("quota")) return "limited";
+  if (source.lastError || status.includes("error") || status.includes("failed")) return "error";
+  if (source.enabled && (source.requiresCredentials ?? false) && !source.hasApiKey) return "missing-config";
+  return "configured";
+}
+
+export function providerUiStateLabel(source: AdminMetadataSourceConfig): string {
+  const state = providerUiState(source);
+  if (state === "missing-config") return "A configurer";
+  if (state === "limited") return "Limite";
+  if (state === "error") return "Erreur";
+  return "Configure";
+}
+
+export function providerUiMessage(source: AdminMetadataSourceConfig): string {
+  const state = providerUiState(source);
+  if (state === "missing-config") return "Source activee, configuration incomplete.";
+  if (state === "limited") return "Quota ou limite temporaire atteint. ReadaBook reessaiera plus tard.";
+  if (state === "error") return "La derniere verification de cette source a echoue.";
+  return source.enabled ? "Source prete." : "Source desactivee.";
+}
+
 export function scheduleSummary(schedule: AutomationScheduleDto, subject: string): string {
   if (schedule.frequency === "disabled") return `${subject} desactive.`;
   if (schedule.frequency === "daily") return `${subject} tous les jours a ${schedule.time}.`;