@koda-sl/baker-cli 0.133.0 → 0.133.3

package/README.md

@@ -3361,7 +3361,7 @@ Ref-image MIMEs: `image/png`, `image/jpeg`, `image/webp`, `image/gif` (via OpenR
 > - **gpt-image** (default frames): clean-plate exclusions placed **LAST** (OpenAI weights early text as composition), an explicit `photorealistic` cue, identity-critical reference **first**, and `quality=high` (gpt-image-2 forces high input fidelity automatically — the `input_fidelity` param is deliberately **not** sent). Gemini keeps exclusions inline; Recraft drops the photoreal cue.
 > - New advisories: `VIDEO_RAW_FACE_KEYFRAME` (a Seedance clip animating a RAW ingested image — real-person 422 risk), `IMAGE_PROMPT_PROFILE_MISSING`, `VIDEO_PROMPT_PROFILE_MISSING`.
 
-> **Content-policy blocks are deterministic, not flaky.** ByteDance's Seedance rejects any first/last frame that reads as a real-person likeness (even an AI-generated face) — surfaced as `content_policy_blocked` (HTTP 422, **non-retryable**). This is ByteDance's upstream filter, so it fires on **any** reseller (Replicate or otherwise) — a provider swap does **not** route around it. Retrying **never** succeeds and wastes credits. Fix the cause: use a **synthetic/AI-generated** (non-identifiable) presenter face, or route a real face to **Veo** (`person_generation: allow_adult`), or make the source frame less photorealistic.
+> **Content-policy blocks are deterministic, not flaky — and now auto-recover.** ByteDance's Seedance rejects any first/last frame that reads as a real-person likeness (even an AI-generated face) — surfaced as `content_policy_blocked` (HTTP 422/`E005` "flagged as sensitive", **non-retryable on the same model**). This is ByteDance's upstream filter, so it fires on **any** reseller (Replicate or otherwise) — a provider swap does **not** route around it. **When a Seedance `video_generate` clip is blocked, the backend automatically re-renders that same clip on `google/veo-3.1-fast` (`person_generation: allow_adult`) — the one curated model that clears the filter — so the run still completes.** The Seedance attempt is never billed (its output was never finalized); only the Veo render is charged. To skip that wasted first attempt entirely (and control look/cost), scaffold a photoreal-cast ad directly on Veo up front (`--real-face`, which also covers AI-generated photoreal faces), or make the source frame less photorealistic.
 
 ---
 
@@ -3966,7 +3966,7 @@ baker canvas run ./reference-ad.video.canvas.json
 | `--select-model <id>` | `~google/gemini-flash-latest` | Override the element-selection `text_generate` model. |
 | `--image-model <id>` | `openai/gpt-5.4-image-2` | Override the per-frame `image_generate` model (defaults to the strongest, matching `scaffold-static-ad`). |
 | `--video-model <id>` | scored router | Override the `video_generate` model **and skip the router**. Curated roster: `bytedance/seedance-2.0` (workhorse), `google/veo-3.1` (cine ceiling + real-face), `google/veo-3.1-fast` (cheap Veo), `kwaivgi/kling-v3.0-pro` (motion-transfer). |
-| `--real-face` | off | Brief needs a real human likeness → the router picks Veo (dodges the ByteDance real-person 422). |
+| `--real-face` | auto-detected | Force the Veo route for a photoreal human presenter (dodges the ByteDance real-person filter). **Usually unnecessary** — the scaffold auto-detects a photoreal on-camera person/animal cast from the deconstruct and routes it to Veo on its own; set this only to force it when the cast isn't detected. Covers AI-generated photoreal faces too, not just real ones. |
 | `--motion-transfer` | off | Motion driven from a reference / hyper-dynamic → the router picks Kling. |
 | `--identity` | off | Same character/product across clips → the router picks the Seedance workhorse. |
 | `--budget <tier>` | `standard` | Router cost posture: `economy` (cheap Veo-fast tier) / `standard` / `premium`. |
@@ -3975,7 +3975,10 @@ baker canvas run ./reference-ad.video.canvas.json
 **Scored model router.** When `--video-model` is omitted, a small weighted scorer picks from the curated
 roster and prints the choice as `models.video_route_reason` in the run report (mirroring Higgsfield's
 `models_explore`). Real-face → Veo, motion-transfer → Kling, else the Seedance workhorse; `--budget economy`
-shifts a Veo pick to the fast tier. The scaffolder also bakes **per-intent param recipes** so the hero/reveal
+shifts a Veo pick to the fast tier. **The router runs after the deconstruct, so a photoreal on-camera
+person/animal cast is auto-detected and routed to Veo up front** — that cast would be blocked on Seedance by
+ByteDance's real-person filter, so picking Seedance would only burn a wasted attempt before the runtime
+auto-fallback re-renders it on Veo anyway. The scaffolder also bakes **per-intent param recipes** so the hero/reveal
 beat claims the 1080p cine ceiling automatically and a Kling hook beat maxes prompt adherence.
 
 Each scene is captured in a **shoot mode** — `ugc_selfie` (talking heads, the default look), `ugc_broll`, `studio_product` (pack shot), `lifestyle_cinematic`, or `screen_ui`. The scaffold derives one per scene (UGC by default; the cinematic and screen lanes are opt-in) and bakes its capture block into the frame and a camera default into the clip; override per scene with a `shoot_mode` field in that scene's `scenes/sNN.json`. Capture aesthetic + depth-of-field follow the mode (UGC stays flat; studio/lifestyle allow shallow DoF). Clips also carry **diegetic native audio** — the scene's own ambience described in the Seedance prompt, never music (the music bed is a separate, ducked track); set a scene's `ambient` field to steer it.

package/dist/{chunk-UBFU3E7L.js → chunk-CIF62V7L.js}

@@ -2488,6 +2488,11 @@ import { readFile as readFile2 } from "fs/promises";
 import path3 from "path";
 import { z as z2 } from "zod";
 
+// src/engine/lib/url.ts
+function looksLikeHttpUrl(value) {
+  return typeof value === "string" && /^https?:\/\//i.test(value.trim());
+}
+
 // src/engine/scaffold/lib/prompt-profiles.ts
 var VEO_PERSON_GENERATION = "allow_adult";
 var VEO_NEGATIVE_PROMPT = "subtitles, captions, on-screen text, watermark, logo, warped face, distorted hands, extra fingers, low quality";
@@ -2720,7 +2725,8 @@ var STAGE_CODES = {
   ODD_DIMENSIONS: "VIDEO_ODD_DIMENSIONS",
   REGION_DROPPED: "VIDEO_REGION_DROPPED",
   OVERLAY_OUT_OF_BOUNDS: "VIDEO_OVERLAY_OUT_OF_BOUNDS",
-  ORPHAN_NODE: "ORPHAN_NODE"
+  ORPHAN_NODE: "ORPHAN_NODE",
+  INGEST_PATH_URL_HEALED: "INGEST_PATH_URL_HEALED"
 };
 var SPAN_MODEL_SLACK_S = 0.25;
 var VIDEO_TIME_SLACK_S = 0.75;
@@ -2777,6 +2783,7 @@ function validateCanvas(input, registry) {
   checkUnknownNodeTypes(ctx);
   if (issues.some((i) => i.code === STAGE_CODES.UNKNOWN_NODE)) return { ok: false, issues };
   checkAllParams(ctx);
+  checkIngestPathUrlHeals(ctx);
   checkAllModelParams(ctx);
   checkAllRefs(ctx);
   checkAllModelInputs(ctx);
@@ -2899,6 +2906,22 @@ function checkAllParams(ctx) {
     pushZodIssues(ctx.issues, parsed.error, `nodes[${i}].params`, STAGE_CODES.PARAMS, n.id, n.type);
   }
 }
+function checkIngestPathUrlHeals(ctx) {
+  for (let i = 0; i < ctx.canvas.nodes.length; i++) {
+    const n = ctx.canvas.nodes[i];
+    if (!n || n.type !== "ingest") continue;
+    const params = n.params ?? {};
+    if (params.source !== "path" || !looksLikeHttpUrl(params.path)) continue;
+    ctx.issues.push({
+      path: `nodes[${i}].params.source`,
+      code: STAGE_CODES.INGEST_PATH_URL_HEALED,
+      message: `ingest "${n.id}" declares source:"path" but its value is a URL \u2014 running it as source:"url". Update the canvas JSON to source:"url" so it matches what baker runs.`,
+      severity: "warning",
+      node_id: n.id,
+      node_type: n.type
+    });
+  }
+}
 function checkAllModelParams(ctx) {
   for (let i = 0; i < ctx.canvas.nodes.length; i++) {
     const n = ctx.canvas.nodes[i];
@@ -4534,18 +4557,30 @@ function mapClientError(ctx, e) {
 // src/engine/nodes/ingest.ts
 var execFile = promisify(execFileCb);
 var ExpectEnum = z5.enum(["image", "video", "audio", "text", "json", "font"]);
-var IngestParams = z5.discriminatedUnion("source", [
-  z5.object({
-    source: z5.literal("url"),
-    url: z5.string().url(),
-    expect: ExpectEnum
-  }).strict(),
-  z5.object({
-    source: z5.literal("path"),
-    path: z5.string().min(1),
-    expect: ExpectEnum
-  }).strict()
-]);
+function coercePathUrlToUrl(input) {
+  if (!input || typeof input !== "object") return input;
+  const obj = input;
+  if (obj.source === "path" && looksLikeHttpUrl(obj.path)) {
+    const { path: _path, ...rest } = obj;
+    return { ...rest, source: "url", url: obj.path.trim() };
+  }
+  return input;
+}
+var IngestParams = z5.preprocess(
+  coercePathUrlToUrl,
+  z5.discriminatedUnion("source", [
+    z5.object({
+      source: z5.literal("url"),
+      url: z5.string().url(),
+      expect: ExpectEnum
+    }).strict(),
+    z5.object({
+      source: z5.literal("path"),
+      path: z5.string().min(1),
+      expect: ExpectEnum
+    }).strict()
+  ])
+);
 var IngestInputs = z5.object({}).strict();
 var IngestOutputs = z5.object({ asset: AssetRef }).strict();
 var RAW_TEXT_EXTENSIONS = [".txt", ".md"];
@@ -7272,8 +7307,8 @@ var videoGenerateNode = delegated({
   id: "video_generate",
   version: "2.0.0",
   category: "video",
-  summary: "Generate video for ad creatives. Curated roster: `bytedance/seedance-2.0` (identity/product workhorse), `google/veo-3.1` (photoreal cine ceiling + real-face fallback), `google/veo-3.1-fast` (cheap Veo iteration), `kwaivgi/kling-3.0` (motion-transfer/dynamic). Async with polling.",
-  when_to_use: "Use `bytedance/seedance-2.0` for identity/product output. Route real human likenesses to `google/veo-3.1` (dodges the ByteDance real-person filter); use `google/veo-3.1-fast` while iterating to keep cost low; `kwaivgi/kling-3.0` for motion-transfer/hyper-dynamic beats. The scaffolder's scored router picks for you. Each model gates its own durations/resolutions/aspect ratios in the registry \u2014 see the README per-model section.",
+  summary: "Generate video for ad creatives. Curated roster: `bytedance/seedance-2.0` (identity/product workhorse, NO photoreal human faces), `google/veo-3.1` (photoreal cine ceiling + the human-face model), `google/veo-3.1-fast` (cheap Veo iteration), `kwaivgi/kling-3.0` (motion-transfer/dynamic). Async with polling.",
+  when_to_use: "Use `bytedance/seedance-2.0` for identity/product output WITHOUT a photoreal human face on camera. Any clip whose frame shows a photoreal human face \u2014 a UGC creator, presenter, or testimonial speaker, REAL OR AI-GENERATED \u2014 must use `google/veo-3.1`: ByteDance's filter rejects photoreal human faces on Seedance (content_policy_blocked/E005), and a synthetic AI face does NOT clear it. Use `google/veo-3.1-fast` while iterating to keep cost low; `kwaivgi/kling-3.0` for motion-transfer/hyper-dynamic beats. `baker canvas scaffold-video` auto-routes a photoreal cast to Veo; a Seedance clip blocked at runtime also auto-recovers on Veo. Each model gates its own durations/resolutions/aspect ratios in the registry \u2014 see the README per-model section.",
   inputs: z28.object({
     first_frame: ImageRef.optional(),
     last_frame: ImageRef.optional(),
@@ -7744,6 +7779,7 @@ export {
   REF_PREFIX,
   parseRefExpr,
   sha256Hex,
+  looksLikeHttpUrl,
   SEEDANCE_PROFILE,
   clipProfileFor,
   clipParamRecipe,
@@ -7762,4 +7798,4 @@ export {
   defaultRegistry,
   createEngineFromEnv
 };
-//# sourceMappingURL=chunk-UBFU3E7L.js.map
+//# sourceMappingURL=chunk-CIF62V7L.js.map