@koda-sl/baker-cli 0.199.0 → 0.200.0-dev.0a9adf6f7

package/dist/cli.js

@@ -18,6 +18,7 @@ import {
   SEEDANCE_PROFILE,
   VIDEO_GENERATE_MODELS,
   ValidationError,
+  YtDlpError,
   clipParamRecipe,
   clipProfileFor,
   collectAssetRefLikes,
@@ -41,6 +42,7 @@ import {
   promptMaxLength,
   requireCredentialsFromEnv,
   resolveConcurrency,
+  runYtDlp,
   sha256Hex,
   spineInputFlags,
   spineInputOps,
@@ -50,7 +52,7 @@ import {
   toModelSafeImage,
   ulid,
   validateCanvasDeep
-} from "./chunk-5MPIOGRO.js";
+} from "./chunk-ONFW7HIK.js";
 import {
   csvOrJson,
   daysAgoIso,
@@ -99,7 +101,7 @@ import {
 } from "./chunk-6NZG2TCM.js";
 
 // src/cli.ts
-import { defineCommand as defineCommand216, runMain } from "citty";
+import { defineCommand as defineCommand217, runMain } from "citty";
 
 // src/commands/actions/index.ts
 import { defineCommand as defineCommand18 } from "citty";
@@ -5686,6 +5688,7 @@ var testimonialsOutscraperWebhookResponseSchema = z18.object({
 // ../api/src/videos.ts
 import { z as z19 } from "zod";
 var videoStatusSchema = z19.enum(["uploading", "uploaded", "processing", "ready", "error"]);
+var videoSourceSchema = z19.enum(["uploaded", "instagram", "url", "youtube"]);
 var videoTranscriptSegmentSchema = z19.object({
   text: z19.string(),
   startSecond: z19.number(),
@@ -5786,6 +5789,19 @@ var videosUploadRequestSchema = z19.object({
   descriptionContext: z19.string().optional()
 });
 var videosUploadResponseSchema = z19.object({ uploadUrl: z19.string(), videoId: z19.string() });
+var videosIngestRequestSchema = z19.object({
+  // Must be a direct media file Mux can pull (mp4/mov/webm/…). A page URL —
+  // a YouTube watch link, a TikTok post — is not one: the CLI resolves those to
+  // a media file before calling this route.
+  url: z19.string().url(),
+  source: videoSourceSchema,
+  externalId: z19.string().optional(),
+  externalUrl: z19.string().optional()
+});
+var videosIngestResponseSchema = z19.object({
+  videoId: z19.string(),
+  deduped: z19.boolean()
+});
 var videosDeleteRequestSchema = z19.object({ id: z19.string().min(1, "Missing video ID") });
 var videosDeleteResponseSchema = z19.object({ ok: z19.literal(true) });
 
@@ -8746,11 +8762,11 @@ function rawTextEntries(value) {
   const values = Array.isArray(value) ? value : typeof value === "string" ? [value] : [];
   return values.filter((v) => typeof v === "string").flatMap((v) => v.split(",")).map((v) => v.trim()).filter(Boolean);
 }
-function rawFileEntries(path35) {
-  if (typeof path35 !== "string" || path35.length === 0) {
+function rawFileEntries(path36) {
+  if (typeof path36 !== "string" || path36.length === 0) {
     return [];
   }
-  return readFileSync2(path35, "utf8").split(/\r?\n/).map((line) => line.trim()).filter((line) => line.length > 0 && !line.startsWith("#"));
+  return readFileSync2(path36, "utf8").split(/\r?\n/).map((line) => line.trim()).filter((line) => line.length > 0 && !line.startsWith("#"));
 }
 function keywordEntries(args) {
   const defaultMatch = typeof args["match-type"] === "string" ? args["match-type"].toUpperCase() : void 0;
@@ -8773,19 +8789,19 @@ function keywordEntries(args) {
   }
   return entries;
 }
-function loadJsonFileArg(path35) {
-  if (typeof path35 !== "string" || path35.length === 0) {
+function loadJsonFileArg(path36) {
+  if (typeof path36 !== "string" || path36.length === 0) {
     return {};
   }
   try {
-    const parsed = JSON.parse(readFileSync2(path35, "utf8"));
+    const parsed = JSON.parse(readFileSync2(path36, "utf8"));
     if (parsed === null || typeof parsed !== "object" || Array.isArray(parsed)) {
-      failWriteValidation(`${path35} must contain a JSON object`);
+      failWriteValidation(`${path36} must contain a JSON object`);
     }
     return parsed;
   } catch (err) {
     if (err instanceof SyntaxError) {
-      failWriteValidation(`${path35} is not valid JSON: ${err.message}`);
+      failWriteValidation(`${path36} is not valid JSON: ${err.message}`);
     }
     throw err;
   }
@@ -8915,10 +8931,10 @@ async function stageUpdate(kind, customerId, target, payload, hints) {
 async function stageTarget(kind, customerId, target, hints) {
   await stageGoogleOp({ kind, customerId, target }, hints);
 }
-async function draftAction(path35, body, chat) {
+async function draftAction(path36, body, chat) {
   try {
     const chatId = resolveChatId(chat);
-    const response = await apiPost(path35, { chatId, ...body });
+    const response = await apiPost(path36, { chatId, ...body });
     writeJsonEnvelope(response);
   } catch (err) {
     handleGoogleError(err);
@@ -11776,19 +11792,19 @@ function failWriteValidation2(message) {
   writeJsonEnvelope({ ok: false, error: { code: "VALIDATION_ERROR", message } });
   process.exit(1);
 }
-function loadJsonFileArg2(path35) {
-  if (typeof path35 !== "string" || path35.length === 0) {
+function loadJsonFileArg2(path36) {
+  if (typeof path36 !== "string" || path36.length === 0) {
     return {};
   }
   try {
-    const parsed = JSON.parse(readFileSync4(path35, "utf8"));
+    const parsed = JSON.parse(readFileSync4(path36, "utf8"));
     if (parsed === null || typeof parsed !== "object" || Array.isArray(parsed)) {
-      failWriteValidation2(`${path35} must contain a JSON object`);
+      failWriteValidation2(`${path36} must contain a JSON object`);
     }
     return parsed;
   } catch (err) {
     if (err instanceof SyntaxError) {
-      failWriteValidation2(`${path35} is not valid JSON: ${err.message}`);
+      failWriteValidation2(`${path36} is not valid JSON: ${err.message}`);
     }
     throw err;
   }
@@ -11873,15 +11889,15 @@ function parseLocaleFlag(value) {
   }
   return { language: match[1], country: match[2].toUpperCase() };
 }
-function loadTargetingFileArg(path35) {
-  if (typeof path35 !== "string" || path35.length === 0) {
+function loadTargetingFileArg(path36) {
+  if (typeof path36 !== "string" || path36.length === 0) {
     return void 0;
   }
-  const parsed = loadJsonFileArg2(path35);
+  const parsed = loadJsonFileArg2(path36);
   const criteria = parsed.targetingCriteria ?? parsed;
   if (!criteria.include) {
     failWriteValidation2(
-      `${path35} must contain targeting criteria with an "include" block (see baker schema ads.linkedin.campaigns.create)`
+      `${path36} must contain targeting criteria with an "include" block (see baker schema ads.linkedin.campaigns.create)`
     );
   }
   return criteria;
@@ -11916,14 +11932,14 @@ function parseCsvLine(line) {
   cells.push(current);
   return cells.map((cell) => cell.trim());
 }
-function parseListFileArg(path35, maxRows) {
-  if (typeof path35 !== "string" || path35.length === 0) {
+function parseListFileArg(path36, maxRows) {
+  if (typeof path36 !== "string" || path36.length === 0) {
     return void 0;
   }
-  const raw = readFileSync4(path35, "utf8");
+  const raw = readFileSync4(path36, "utf8");
   const lines = raw.split(/\r?\n/).filter((line) => line.trim().length > 0);
   if (lines.length < 2) {
-    failWriteValidation2(`${path35} needs a header row and at least one data row`);
+    failWriteValidation2(`${path36} needs a header row and at least one data row`);
   }
   const columns = parseCsvLine(lines[0]).map((column) => column.trim());
   const rows = [];
@@ -11942,7 +11958,7 @@ function parseListFileArg(path35, maxRows) {
     }
   }
   if (rows.length > maxRows) {
-    failWriteValidation2(`${path35} has ${rows.length} rows \u2014 the inline limit is ${maxRows}. Split the list.`);
+    failWriteValidation2(`${path36} has ${rows.length} rows \u2014 the inline limit is ${maxRows}. Split the list.`);
   }
   return { columns, rows };
 }
@@ -12038,11 +12054,11 @@ function readPositionals(args) {
 function splitIdList(raw) {
   return raw.split(",").map((id) => id.trim()).filter(Boolean);
 }
-function idsFileEntries(path35) {
-  if (typeof path35 !== "string" || path35.length === 0) {
+function idsFileEntries(path36) {
+  if (typeof path36 !== "string" || path36.length === 0) {
     return [];
   }
-  return readFileSync4(path35, "utf8").split(/\r?\n/).map((line) => line.trim()).filter((line) => line.length > 0 && !line.startsWith("#")).flatMap(splitIdList);
+  return readFileSync4(path36, "utf8").split(/\r?\n/).map((line) => line.trim()).filter((line) => line.length > 0 && !line.startsWith("#")).flatMap(splitIdList);
 }
 function requireTargets(args, entity) {
   const positionals = readPositionals(args);
@@ -14665,9 +14681,9 @@ function compactRow(row) {
     ...destination.postUrn ? { postUrn: destination.postUrn } : {}
   };
 }
-function readPath(row, path35) {
+function readPath(row, path36) {
   let current = row;
-  for (const segment of path35.split(".")) {
+  for (const segment of path36.split(".")) {
     const record = asRecord2(current);
     if (!record) return void 0;
     current = record[segment];
@@ -14677,10 +14693,10 @@ function readPath(row, path35) {
 function projectFields(rows, paths) {
   return rows.map((row) => {
     const projected = {};
-    for (const path35 of paths) {
-      const value = readPath(row, path35);
+    for (const path36 of paths) {
+      const value = readPath(row, path36);
       if (value !== void 0) {
-        projected[path35] = value;
+        projected[path36] = value;
       }
     }
     return projected;
@@ -15974,11 +15990,11 @@ var updateStatusSchema = z21.enum(UPDATE_STATUSES);
 function currencyMinimums2(currencyCode) {
   return CURRENCY_MINIMUMS2[currencyCode] ?? DEFAULT_CURRENCY_MINIMUM2;
 }
-function validateDailyBudgetFloor(money, ctx, path35) {
+function validateDailyBudgetFloor(money, ctx, path36) {
   if (money?.currencyCode) {
     const min = currencyMinimums2(money.currencyCode).dailyBudgetMin;
     if (Number(money.amount) < min) {
-      ctx.addIssue({ code: "custom", path: path35, message: `below the ${min} ${money.currencyCode} daily minimum` });
+      ctx.addIssue({ code: "custom", path: path36, message: `below the ${min} ${money.currencyCode} daily minimum` });
     }
   }
 }
@@ -16640,19 +16656,19 @@ function failWriteValidation3(message) {
   writeJsonEnvelope({ ok: false, error: { code: "VALIDATION_ERROR", message } });
   process.exit(1);
 }
-function loadJsonFileArg3(path35) {
-  if (typeof path35 !== "string" || path35.length === 0) {
+function loadJsonFileArg3(path36) {
+  if (typeof path36 !== "string" || path36.length === 0) {
     return {};
   }
   try {
-    const parsed = JSON.parse(readFileSync8(path35, "utf8"));
+    const parsed = JSON.parse(readFileSync8(path36, "utf8"));
     if (parsed === null || typeof parsed !== "object" || Array.isArray(parsed)) {
-      failWriteValidation3(`${path35} must contain a JSON object`);
+      failWriteValidation3(`${path36} must contain a JSON object`);
     }
     return parsed;
   } catch (err) {
     if (err instanceof SyntaxError) {
-      failWriteValidation3(`${path35} is not valid JSON: ${err.message}`);
+      failWriteValidation3(`${path36} is not valid JSON: ${err.message}`);
     }
     throw err;
   }
@@ -27725,12 +27741,12 @@ function listFlowSlugs() {
   return readdirSync2(dir, { withFileTypes: true }).filter((entry) => entry.isDirectory() && !entry.name.startsWith("_") && entry.name !== ".gitkeep").map((entry) => entry.name).sort();
 }
 function readFlowTree(slug) {
-  const path35 = join3(flowsDir(), slug, "_data.json");
-  if (!existsSync4(path35)) {
+  const path36 = join3(flowsDir(), slug, "_data.json");
+  if (!existsSync4(path36)) {
     failLocal(`No form "${slug}". Run "baker flows list" to see the forms in this workspace.`);
   }
   try {
-    return JSON.parse(readFileSync9(path35, "utf-8"));
+    return JSON.parse(readFileSync9(path36, "utf-8"));
   } catch (error) {
     failLocal(`Could not read form "${slug}": ${error instanceof Error ? error.message : String(error)}`);
   }
@@ -28143,10 +28159,10 @@ async function stageOps(ops) {
     handleError(err);
   }
 }
-async function draftAction2(path35, body, chat) {
+async function draftAction2(path36, body, chat) {
   const chatId = resolveChatId(chat);
   try {
-    const data = await apiPost(path35, { chatId, ...body });
+    const data = await apiPost(path36, { chatId, ...body });
     writeJsonEnvelope({ ok: true, data });
     return data;
   } catch (err) {
@@ -30167,9 +30183,9 @@ async function readImageBuffer(pathOrUrl) {
   }
   return readFile19(pathOrUrl);
 }
-async function isDirectory(path35) {
+async function isDirectory(path36) {
   try {
-    const s = await stat4(path35);
+    const s = await stat4(path36);
     return s.isDirectory();
   } catch {
     return false;
@@ -30453,13 +30469,13 @@ function resolveDownloadPath({ baseName, extension, out, outIsDirectory: outIsDi
 }
 function disambiguate(paths) {
   const taken = /* @__PURE__ */ new Set();
-  return paths.map((path35) => {
-    if (!taken.has(path35)) {
-      taken.add(path35);
-      return path35;
+  return paths.map((path36) => {
+    if (!taken.has(path36)) {
+      taken.add(path36);
+      return path36;
     }
-    const ext = extname3(path35);
-    const stem = path35.slice(0, path35.length - ext.length);
+    const ext = extname3(path36);
+    const stem = path36.slice(0, path36.length - ext.length);
     let n = 2;
     while (taken.has(`${stem}-${n}${ext}`)) n += 1;
     const unique = `${stem}-${n}${ext}`;
@@ -30586,10 +30602,10 @@ async function runDownloads(plan) {
   const paths = disambiguate(fetched.map((item) => item.path));
   const downloaded = [];
   for (const [index, item] of fetched.entries()) {
-    const path35 = paths[index] ?? item.path;
+    const path36 = paths[index] ?? item.path;
     try {
-      await atomicWrite(path35, item.buffer);
-      downloaded.push({ input: item.input, output: path35, bytes: item.buffer.length, contentType: item.contentType });
+      await atomicWrite(path36, item.buffer);
+      downloaded.push({ input: item.input, output: path36, bytes: item.buffer.length, contentType: item.contentType });
     } catch (err) {
       failed.push({ input: item.input, error: failureMessage(err, "Write failed") });
     }
@@ -40300,7 +40316,7 @@ function parseImageRefs(spec) {
 }
 var defaultDeps = {
   ingest: (url) => apiPost("/api/images/ingest", { url, source: "uploaded" }),
-  upload: (path35) => uploadLocalImage({ file: path35, contentType: detectImageContentType(path35), source: "uploaded" })
+  upload: (path36) => uploadLocalImage({ file: path36, contentType: detectImageContentType(path36), source: "uploaded" })
 };
 async function resolveLibraryImageIds(spec, limit, deps = defaultDeps) {
   const refs = parseImageRefs(spec);
@@ -40320,10 +40336,10 @@ async function resolveLibraryImageIds(spec, limit, deps = defaultDeps) {
   }
   return { imageIds, added };
 }
-function uploadFailure(path35) {
+function uploadFailure(path36) {
   return (error) => {
     if (error instanceof ApiError) throw error;
-    throw new ApiError("VALIDATION_ERROR", `Could not read "${path35}" as an image.`);
+    throw new ApiError("VALIDATION_ERROR", `Could not read "${path36}" as an image.`);
   };
 }
 
@@ -41373,10 +41389,10 @@ async function stageOp4(op) {
     handleError4(err);
   }
 }
-async function draftAction3(path35, body, chat) {
+async function draftAction3(path36, body, chat) {
   const chatId = resolveChatId(chat);
   try {
-    const data = await apiPost(path35, { chatId, ...body });
+    const data = await apiPost(path36, { chatId, ...body });
     writeJsonEnvelope({ ok: true, data });
     return data;
   } catch (err) {
@@ -42292,7 +42308,7 @@ Full guide: __tooling__/docs/tools/baker/testimonials.md`
 });
 
 // src/commands/videos/index.ts
-import { defineCommand as defineCommand202 } from "citty";
+import { defineCommand as defineCommand203 } from "citty";
 
 // src/commands/videos/delete.ts
 import { defineCommand as defineCommand197 } from "citty";
@@ -42423,8 +42439,214 @@ var groupCommand2 = defineCommand199({
   }
 });
 
-// src/commands/videos/search.ts
+// src/commands/videos/ingest.ts
+import { mkdtemp as mkdtemp2, readFile as readFile24, rm as rm7, stat as stat7 } from "fs/promises";
+import { tmpdir as tmpdir3 } from "os";
+import path35 from "path";
 import { defineCommand as defineCommand200 } from "citty";
+
+// src/commands/videos/ingestUrl.ts
+var MAX_VIDEO_INGEST_BYTES = 2 * 1024 * 1024 * 1024;
+var INGEST_MAX_HEIGHT = 1080;
+var DIRECT_MEDIA_EXTENSIONS = [".mp4", ".mov", ".webm", ".m4v", ".mkv", ".avi"];
+var YOUTUBE_HOSTS = ["youtube.com", "youtu.be", "m.youtube.com", "www.youtube.com"];
+function isDirectMediaUrl(rawUrl) {
+  const pathname = safePathname(rawUrl);
+  return DIRECT_MEDIA_EXTENSIONS.some((ext) => pathname.endsWith(ext));
+}
+function resolveIngestSource(explicit, rawUrl) {
+  if (explicit) return explicit;
+  return isYouTubeUrl(rawUrl) ? "youtube" : "url";
+}
+function isYouTubeUrl(rawUrl) {
+  let host;
+  try {
+    host = new URL(rawUrl).hostname.toLowerCase();
+  } catch {
+    return false;
+  }
+  return YOUTUBE_HOSTS.includes(host);
+}
+var TOOLING_OR_BLOCK_SIGNATURES = [
+  "requested format is not available",
+  "only images are available",
+  "nsig",
+  "sign in to confirm",
+  // bot check — distinct from "Private video. Sign in if…"
+  "unable to download api page",
+  "http error 403"
+];
+function isToolingOrBlockFailure(detail) {
+  const haystack = detail.toLowerCase();
+  return TOOLING_OR_BLOCK_SIGNATURES.some((signature) => haystack.includes(signature));
+}
+function safePathname(rawUrl) {
+  try {
+    return new URL(rawUrl).pathname.toLowerCase();
+  } catch {
+    return "";
+  }
+}
+
+// src/commands/videos/ingest.ts
+registerSchema({
+  command: "videos.ingest",
+  description: "Add a video to the library from a URL \u2014 a direct file, or a YouTube/TikTok/Vimeo page.",
+  args: {
+    url: { type: "string", description: "Video URL (direct file or a watch/post page)", required: true },
+    source: { type: "string", description: "Provenance: url | youtube", required: false },
+    "external-id": { type: "string", description: "Provider asset id (deduped on)", required: false },
+    "external-url": { type: "string", description: "Canonical page URL", required: false },
+    download: {
+      type: "boolean",
+      description: "Force the download path instead of letting Baker fetch the URL directly",
+      required: false,
+      default: false
+    },
+    "dry-run": { type: "boolean", description: "Preview the operation without executing", required: false }
+  }
+});
+var ingestCommand2 = defineCommand200({
+  meta: {
+    name: "ingest",
+    description: "Add a video to the library from a URL. A direct file URL is handed straight to Baker, which fetches it. A page URL (YouTube, TikTok, Vimeo, Instagram) is downloaded here first, then uploaded \u2014 and a direct URL that Baker cannot fetch falls back to that same path automatically.\n\nExample: baker videos ingest https://www.youtube.com/watch?v=abc123"
+  },
+  args: {
+    url: { type: "positional", description: "Video URL", required: false },
+    source: { type: "string", description: "Provenance: url | youtube", required: false },
+    "external-id": { type: "string", description: "Provider asset id", required: false },
+    "external-url": { type: "string", description: "Canonical page URL", required: false },
+    download: { type: "boolean", description: "Force the download path", required: false, default: false },
+    "dry-run": { type: "boolean", description: "Preview without executing", required: false, default: false }
+  },
+  run: ({ args }) => runVideoIngest({
+    url: args.url,
+    source: args.source,
+    externalId: args["external-id"],
+    externalUrl: args["external-url"],
+    forceDownload: args.download === true,
+    dryRun: args["dry-run"] === true
+  })
+});
+async function runVideoIngest(opts) {
+  const url = opts.url;
+  if (!url) {
+    writeJson({ ok: false, error: { code: "VALIDATION_ERROR", message: "URL is required" } });
+    process.exit(1);
+  }
+  const direct = !opts.forceDownload && isDirectMediaUrl(url);
+  const args = {
+    url,
+    source: resolveIngestSource(opts.source, url),
+    externalId: opts.externalId,
+    externalUrl: opts.externalUrl ?? (direct ? void 0 : url)
+  };
+  if (opts.dryRun) {
+    writeJson({
+      ok: true,
+      dryRun: true,
+      operation: "videos.ingest",
+      params: {
+        url,
+        source: args.source,
+        externalId: args.externalId ?? null,
+        externalUrl: args.externalUrl ?? null,
+        direct
+      }
+    });
+    return;
+  }
+  try {
+    writeJson({ ok: true, data: await ingestByRoute(args, direct) });
+  } catch (err) {
+    reportIngestFailure(err);
+  }
+}
+async function ingestByRoute(args, direct) {
+  if (!direct) {
+    return { ...await downloadThenIngest(args), via: "download" };
+  }
+  try {
+    return { ...await ingestUrl(args), via: "direct" };
+  } catch (err) {
+    if (!(err instanceof ApiError)) throw err;
+    return { ...await downloadThenIngest(args), via: "download", fallbackFrom: err.message };
+  }
+}
+function reportIngestFailure(err) {
+  if (err instanceof ApiError) {
+    writeJson({ ok: false, error: { code: err.code, message: err.message } });
+    process.exit(1);
+  }
+  if (err instanceof YtDlpError) {
+    const toolingOrBlock = isToolingOrBlockFailure(`${err.stderrTail} ${err.message}`);
+    writeJson({
+      ok: false,
+      error: {
+        code: "INGEST_FAILED",
+        message: `Couldn't download that video. ${err.stderrTail || err.message}`,
+        fix: toolingOrBlock ? "The site wouldn't hand over the video \u2014 usually the downloader being out of date against a site that changed, or the request being blocked. Retrying won't help, and the link is probably fine. Report it so the tool can be updated; meanwhile download the file and use `baker videos upload <file>`." : "Check the link is public and playable. Private, age-restricted, or region-locked videos can't be downloaded \u2014 download the file yourself and use `baker videos upload <file>`."
+      }
+    });
+    process.exit(1);
+  }
+  writeJson({ ok: false, error: { code: "INTERNAL_ERROR", message: "Unexpected error" } });
+  process.exit(1);
+}
+function ingestUrl(args) {
+  const body = {
+    url: args.url,
+    source: args.source
+  };
+  if (args.externalId) body.externalId = args.externalId;
+  if (args.externalUrl) body.externalUrl = args.externalUrl;
+  return apiPost("/api/videos/ingest", body);
+}
+async function downloadThenIngest(args) {
+  const workDir = await mkdtemp2(path35.join(tmpdir3(), "videos-ingest-"));
+  try {
+    const { filePath, info } = await runYtDlp({
+      url: args.url,
+      audioOnly: false,
+      workDir,
+      maxHeight: INGEST_MAX_HEIGHT
+    });
+    const stats = await stat7(filePath);
+    if (stats.size > MAX_VIDEO_INGEST_BYTES) {
+      throw new ApiError(
+        "VALIDATION_ERROR",
+        `That video is ${Math.round(stats.size / 1e6)} MB, over the ${Math.round(MAX_VIDEO_INGEST_BYTES / 1e6)} MB limit for this route.`
+      );
+    }
+    const bytes = await readFile24(filePath);
+    const publicUrl = await uploadToAssetStore(bytes);
+    const externalId = args.externalId ?? (typeof info.id === "string" ? `${args.source}:${info.id}` : void 0);
+    return await ingestUrl({ ...args, url: publicUrl, externalId, externalUrl: args.externalUrl ?? args.url });
+  } finally {
+    await rm7(workDir, { recursive: true, force: true }).catch(() => {
+    });
+  }
+}
+async function uploadToAssetStore(bytes) {
+  const sha256 = sha256Hex(bytes);
+  const { putUrl, publicUrl } = await apiPost("/api/canvas/assets/presign", {
+    sha256,
+    mime: "video/mp4",
+    purpose: "source"
+  });
+  const putRes = await fetch(putUrl, {
+    method: "PUT",
+    body: new Uint8Array(bytes),
+    headers: { "Content-Type": "video/mp4" }
+  });
+  if (!putRes.ok) {
+    throw new ApiError("INTERNAL_ERROR", `Storing the downloaded video failed: HTTP ${putRes.status}`);
+  }
+  return publicUrl;
+}
+
+// src/commands/videos/search.ts
+import { defineCommand as defineCommand201 } from "citty";
 registerSchema({
   command: "videos.search",
   description: "Search videos by text query. Only returns ready videos.",
@@ -42434,7 +42656,7 @@ registerSchema({
     tags: { type: "string", description: "Comma-separated tags to filter by", required: false }
   }
 });
-var searchCommand4 = defineCommand200({
+var searchCommand4 = defineCommand201({
   meta: {
     name: "search",
     description: "Semantic search videos by text query. Uses hybrid BM25 + vector + reranking. Example: baker videos search 'product demo' --tags tutorial"
@@ -42484,9 +42706,9 @@ var searchCommand4 = defineCommand200({
 var tagsCommand5 = makeTagsCommand("videos", "video", "/api/videos/tags");
 
 // src/commands/videos/upload.ts
-import { readFile as readFile24, stat as stat7 } from "fs/promises";
+import { readFile as readFile25, stat as stat8 } from "fs/promises";
 import { basename as basename3, extname as extname4 } from "path";
-import { defineCommand as defineCommand201 } from "citty";
+import { defineCommand as defineCommand202 } from "citty";
 var MIME_MAP = {
   ".mp4": "video/mp4",
   ".mov": "video/quicktime",
@@ -42496,9 +42718,9 @@ var MIME_MAP = {
 };
 registerSchema({
   command: "videos.upload",
-  description: "Upload a video file to Baker (via Mux direct upload)",
+  description: "Upload a video to Baker \u2014 a local file (via Mux direct upload) or a remote http(s) URL.",
   args: {
-    file: { type: "string", description: "Path to the video file", required: true },
+    file: { type: "string", description: "Local video file path, or a remote http(s) URL", required: true },
     "content-type": {
       type: "string",
       description: "MIME type (auto-detected from extension if omitted)",
@@ -42525,33 +42747,49 @@ function detectContentType(filePath) {
   }
   return mime;
 }
-var uploadCommand2 = defineCommand201({
+function isRemoteUrl3(value) {
+  return /^https?:\/\//i.test(value);
+}
+var uploadCommand2 = defineCommand202({
   meta: {
     name: "upload",
-    description: "Upload a video file to Baker via Mux direct upload. Auto-detects content type. Example: baker videos upload ./demo.mp4"
+    description: "Upload a video to Baker \u2014 accepts a local file path OR a remote http(s) URL.\n\nLocal: auto-detects content type and uploads via Mux direct upload.\nRemote: hands off to `videos ingest` (direct fetch, or download-then-upload for a YouTube/TikTok/Vimeo page).\n\nExamples:\n  baker videos upload ./demo.mp4\n  baker videos upload https://www.youtube.com/watch?v=abc123"
   },
   args: {
-    file: { type: "positional", description: "Path to the video file", required: false },
-    "content-type": { type: "string", description: "MIME type (auto-detected if omitted)", required: false },
+    file: { type: "positional", description: "Local video file path or remote http(s) URL", required: false },
+    "content-type": { type: "string", description: "MIME type (local only \u2014 auto-detected)", required: false },
     context: {
       type: "string",
       description: "What this clip is and what it is for \u2014 who is on camera, which campaign, how it was shot. Start here: the analysis reads it, so it is what makes the clip findable later.",
       required: false
     },
+    source: { type: "string", description: "Provenance (URL mode only): url | youtube", required: false },
+    "external-id": { type: "string", description: "Provider asset id (URL mode only)", required: false },
+    "external-url": { type: "string", description: "Canonical page URL (URL mode only)", required: false },
     "dry-run": { type: "boolean", description: "Preview without executing", required: false, default: false }
   },
   run: async ({ args }) => {
     try {
       const filePath = args.file;
       if (!filePath) {
-        writeJson({ ok: false, error: { code: "VALIDATION_ERROR", message: "File path is required" } });
+        writeJson({ ok: false, error: { code: "VALIDATION_ERROR", message: "File path or URL is required" } });
         process.exit(1);
       }
+      if (isRemoteUrl3(filePath)) {
+        await runVideoIngest({
+          url: filePath,
+          source: args.source,
+          externalId: args["external-id"],
+          externalUrl: args["external-url"],
+          dryRun: args["dry-run"] === true
+        });
+        return;
+      }
       const contentType = args["content-type"] || detectContentType(filePath);
       const originalFilename = basename3(filePath);
       const descriptionContext = args.context;
       if (args["dry-run"]) {
-        const fileStats = await stat7(filePath);
+        const fileStats = await stat8(filePath);
         writeJson({
           ok: true,
           dryRun: true,
@@ -42564,7 +42802,7 @@ var uploadCommand2 = defineCommand201({
         originalFilename,
         descriptionContext
       });
-      const fileBuffer = await readFile24(filePath);
+      const fileBuffer = await readFile25(filePath);
       const uploadResponse = await fetch(uploadUrl, {
         method: "PUT",
         headers: { "Content-Type": contentType },
@@ -42595,16 +42833,17 @@ var uploadCommand2 = defineCommand201({
 });
 
 // src/commands/videos/index.ts
-var videosCommand = defineCommand202({
+var videosCommand = defineCommand203({
   meta: {
     name: "videos",
-    description: `Find and manage videos in Baker. Subcommands: search, get, upload, delete, tags.
+    description: `Find and manage videos in Baker. Subcommands: search, get, upload, ingest, delete, tags.
 
 Examples:
   baker videos search "product demo" --limit 5
   baker videos search "tutorial" --tags explainer
   baker videos get <video-id>
   baker videos upload ./demo.mp4
+  baker videos ingest https://www.youtube.com/watch?v=abc123
   baker videos delete <video-id> --dry-run
   baker videos tags
 Full guide: __tooling__/docs/tools/baker/videos.md`
@@ -42614,16 +42853,17 @@ Full guide: __tooling__/docs/tools/baker/videos.md`
     group: groupCommand2,
     search: searchCommand4,
     upload: uploadCommand2,
+    ingest: ingestCommand2,
     delete: deleteCommand3,
     tags: tagsCommand5
   }
 });
 
 // src/commands/winning-ads/index.ts
-import { defineCommand as defineCommand215 } from "citty";
+import { defineCommand as defineCommand216 } from "citty";
 
 // src/commands/winning-ads/advertisers.ts
-import { defineCommand as defineCommand203 } from "citty";
+import { defineCommand as defineCommand204 } from "citty";
 
 // src/commands/winning-ads/shared.ts
 function splitList2(value) {
@@ -42676,7 +42916,7 @@ function advertiserNormalizer(record, full) {
     last_synced_at: record.last_synced_at ?? null
   };
 }
-var advertisersCommand2 = defineCommand203({
+var advertisersCommand2 = defineCommand204({
   meta: {
     name: "advertisers",
     description: 'List corpus advertisers by name or domain. Find your own advertiser for --exclude-advertiser, or a competitor for --advertiser-id / winners. Example: baker winning-ads advertisers "Deel" --output md'
@@ -42734,7 +42974,7 @@ var advertisersCommand2 = defineCommand203({
 });
 
 // src/commands/winning-ads/brief.ts
-import { defineCommand as defineCommand204 } from "citty";
+import { defineCommand as defineCommand205 } from "citty";
 registerSchema({
   command: "winning-ads.brief",
   description: "Generate a creative brief grounded in strategically-similar winning ads. Optionally describe the target creative with --dna (JSON) and steer with --notes.",
@@ -42780,7 +43020,7 @@ function parseDna(raw) {
   }
   return parsed;
 }
-var briefCommand = defineCommand204({
+var briefCommand = defineCommand205({
   meta: {
     name: "brief",
     description: `Generate a creative brief from winning references. Example: baker winning-ads brief --dna '{"angle":"cost savings"}' --notes "B2B, LinkedIn video" --k 8`
@@ -42816,7 +43056,7 @@ var briefCommand = defineCommand204({
 });
 
 // src/commands/winning-ads/content.ts
-import { defineCommand as defineCommand205 } from "citty";
+import { defineCommand as defineCommand206 } from "citty";
 registerSchema({
   command: "winning-ads.content",
   description: "Read what's INSIDE one winning ad: the spoken transcript, the on-screen text, and the ad copy. Use this after `search`/`winners`/`feed` return a shortlist \u2014 pass an ad_id to understand a reference before reproducing it. Add --full for speech, pacing, and soundtrack detail. Video ads carry the transcript/on-screen text; static ads carry only the copy.",
@@ -42829,7 +43069,7 @@ registerSchema({
     }
   }
 });
-var contentCommand = defineCommand205({
+var contentCommand = defineCommand206({
   meta: {
     name: "content",
     description: "Read the transcript + on-screen text + copy of one winning ad. Example: baker winning-ads content adg_123 --platform meta --full --output md"
@@ -42878,7 +43118,7 @@ var contentCommand = defineCommand205({
 });
 
 // src/commands/winning-ads/feed.ts
-import { defineCommand as defineCommand206 } from "citty";
+import { defineCommand as defineCommand207 } from "citty";
 function buildFeedParams(input) {
   const params = {};
   const advertiser = splitList2(input.advertiser);
@@ -42930,7 +43170,7 @@ registerSchema({
     format: { type: "string", description: "Comma-separated formats to include (e.g. static,video)", required: false }
   }
 });
-var feedCommand = defineCommand206({
+var feedCommand = defineCommand207({
   meta: {
     name: "feed",
     description: "Winners across every brand you follow (browse, then trim per advertiser). Example: baker winning-ads feed --per-advertiser 5 --output md"
@@ -43015,7 +43255,7 @@ var feedCommand = defineCommand206({
 });
 
 // src/commands/winning-ads/follow.ts
-import { defineCommand as defineCommand207 } from "citty";
+import { defineCommand as defineCommand208 } from "citty";
 var PLATFORMS = ["meta", "linkedin"];
 registerSchema({
   command: "winning-ads.follow",
@@ -43030,7 +43270,7 @@ registerSchema({
     label: { type: "string", description: "Optional display label (defaults to the resolved name)", required: false }
   }
 });
-var followCommand = defineCommand207({
+var followCommand = defineCommand208({
   meta: {
     name: "follow",
     description: 'Follow a brand to track ALL its ads \u2014 every platform and country. --platform is how we read your input, not a limit. A domain tracks both Meta + LinkedIn. Example: baker winning-ads follow "deel.com" --platform meta'
@@ -43077,7 +43317,7 @@ var followCommand = defineCommand207({
 });
 
 // src/commands/winning-ads/follow-competitors.ts
-import { defineCommand as defineCommand208 } from "citty";
+import { defineCommand as defineCommand209 } from "citty";
 var PLATFORMS2 = ["meta", "linkedin"];
 var BATCH_TIMEOUT_MS = 3e5;
 function buildFollowBatchBody(input) {
@@ -43110,7 +43350,7 @@ registerSchema({
     }
   }
 });
-var followCompetitorsCommand = defineCommand208({
+var followCompetitorsCommand = defineCommand209({
   meta: {
     name: "follow-competitors",
     description: 'Follow many brands at once by domain \u2014 add every competitor in one call. Example: baker winning-ads follow-competitors "deel.com,notion.so,hubspot.com"'
@@ -43185,7 +43425,7 @@ var followCompetitorsCommand = defineCommand208({
 });
 
 // src/commands/winning-ads/following.ts
-import { defineCommand as defineCommand209 } from "citty";
+import { defineCommand as defineCommand210 } from "citty";
 registerSchema({
   command: "winning-ads.following",
   description: "List the brands you follow in your ad-dna library, with each one's status (ready vs still adding) and cached ad counts.",
@@ -43218,7 +43458,7 @@ function followingNormalizer(record, full) {
     platforms: Array.isArray(record.platforms) ? record.platforms : []
   };
 }
-var followingCommand = defineCommand209({
+var followingCommand = defineCommand210({
   meta: {
     name: "following",
     description: "List brands you follow, with status (ready / adding\u2026) and cached counts. Example: baker winning-ads following --output md"
@@ -43253,7 +43493,7 @@ var followingCommand = defineCommand209({
 });
 
 // src/commands/winning-ads/patterns.ts
-import { defineCommand as defineCommand210 } from "citty";
+import { defineCommand as defineCommand211 } from "citty";
 registerSchema({
   command: "winning-ads.patterns",
   description: "Mine what separates two cohorts of ads: pass a comma-list of winning ad ids (--winners) and a comma-list of weaker ad ids (--duds). Returns the discriminating DNA fields.",
@@ -43292,7 +43532,7 @@ function discriminatorRow(record) {
     top_values_duds: Array.isArray(record.top_values_b) ? record.top_values_b.join(", ") : ""
   };
 }
-var patternsCommand = defineCommand210({
+var patternsCommand = defineCommand211({
   meta: {
     name: "patterns",
     description: "Discover what separates winning ads from weak ones. Example: baker winning-ads patterns --winners a_1,a_2,a_3 --duds a_9,a_8 --output md"
@@ -43348,7 +43588,7 @@ var patternsCommand = defineCommand210({
 });
 
 // src/commands/winning-ads/search.ts
-import { defineCommand as defineCommand211 } from "citty";
+import { defineCommand as defineCommand212 } from "citty";
 registerSchema({
   command: "winning-ads.search",
   description: "Search the ad-dna corpus of scored winning ads. Returns a lean shortlist (advertiser, summary, scores, media_url) to pick a reference to reproduce.",
@@ -43456,7 +43696,7 @@ function buildSearchBody2(args) {
   }
   return body;
 }
-var searchCommand5 = defineCommand211({
+var searchCommand5 = defineCommand212({
   meta: {
     name: "search",
     description: "Search winning reference ads. Example: baker winning-ads search 'B2B SaaS before/after AI automation' --platform meta --format static --winner-category winner --exclude-advertiser adv_123 --output md"
@@ -43571,7 +43811,7 @@ var searchCommand5 = defineCommand211({
 });
 
 // src/commands/winning-ads/seeds.ts
-import { defineCommand as defineCommand212 } from "citty";
+import { defineCommand as defineCommand213 } from "citty";
 function leanRow(r) {
   return {
     key: r.key,
@@ -43599,7 +43839,7 @@ function makeSeedCommand(opts) {
       limit: { type: "number", description: "Max keys 1-100 (default 20)", required: false, default: 20 }
     }
   });
-  return defineCommand212({
+  return defineCommand213({
     meta: { name: opts.name, description: opts.description },
     args: {
       platform: { type: "string", description: "Single platform to segment on", required: false },
@@ -43648,7 +43888,7 @@ var formatsCommand = makeSeedCommand({
 });
 
 // src/commands/winning-ads/unfollow.ts
-import { defineCommand as defineCommand213 } from "citty";
+import { defineCommand as defineCommand214 } from "citty";
 registerSchema({
   command: "winning-ads.unfollow",
   description: "Stop following a brand \u2014 removes it from your ad-dna library by advertiser id.",
@@ -43656,7 +43896,7 @@ registerSchema({
     advertiser: { type: "string", description: "Advertiser id to unfollow", required: true }
   }
 });
-var unfollowCommand = defineCommand213({
+var unfollowCommand = defineCommand214({
   meta: {
     name: "unfollow",
     description: "Stop following a brand by advertiser id. Example: baker winning-ads unfollow adv_123"
@@ -43677,7 +43917,7 @@ var unfollowCommand = defineCommand213({
 });
 
 // src/commands/winning-ads/winners.ts
-import { defineCommand as defineCommand214 } from "citty";
+import { defineCommand as defineCommand215 } from "citty";
 registerSchema({
   command: "winning-ads.winners",
   description: "Top winning ads for one advertiser id (from `advertisers` or `following`). Returns lean winner cards; add --full for DNA + longevity.",
@@ -43687,7 +43927,7 @@ registerSchema({
     platform: { type: "string", description: "Filter to a single platform: meta|linkedin", required: false }
   }
 });
-var winnersCommand = defineCommand214({
+var winnersCommand = defineCommand215({
   meta: {
     name: "winners",
     description: "Top winning ads for a specific advertiser id. Example: baker winning-ads winners adv_123 --top 15 --output md"
@@ -43737,7 +43977,7 @@ var winnersCommand = defineCommand214({
 });
 
 // src/commands/winning-ads/index.ts
-var winningAdsCommand = defineCommand215({
+var winningAdsCommand = defineCommand216({
   meta: {
     name: "winning-ads",
     description: `Search the ad-dna corpus of scored "winning" ads for reference creatives to reproduce, and manage the brands your library tracks. Proxied through the Baker backend (BAKER_API_KEY) \u2014 no separate token needed.
@@ -43809,7 +44049,7 @@ function getCliVersion() {
 }
 
 // src/cli.ts
-var main = defineCommand216({
+var main = defineCommand217({
   meta: {
     name: "baker",
     version: getCliVersion(),