@starreel/mcp 0.1.218 → 0.1.220

package/SKILL.md

@@ -95,6 +95,7 @@ announces a condensed version as MCP `instructions` at connect time.
 | Their own **footage** for a shot (screen recording, product b-roll, an existing clip) | `upload_shot_footage` → that shot is no longer AI-generated (frames/video skipped, final cut uses the clip as-is, duration written back from the clip); **it overwrites whatever AI video that shot already had**, and afterwards every AI video call on it (including `edit_video_shot`) is refused with **409** until you `clear_shot_footage` or pass `replace_user_footage: true` | pasting an externally AI-generated clip to "fix" a shot — it carries none of this film's identity / style anchors; use `regenerate_shot_video` instead |
 | A clip that should **drive the motion** of an AI shot (a blocking/previz pass, a dance or action reference) | `edit_video_shot` with `reference_video_urls` — the shot keeps its AI video and borrows the clip's movement. To make the clip itself the source and restyle it in place, `upload_shot_footage` it first, then `edit_video_shot` with `replace_user_footage: true` | `upload_shot_footage` alone — that registers the clip as the finished shot, so nothing gets restyled and every later AI call on it returns 409 |
 | A clip whose **motion should be kept while the people / outfits are redrawn** (video repaint) — their own footage or a shot already generated here | `get_repaint_capabilities` (is it enabled for this drama; `relayMaxSegments`) → for their own clip, `upload_shot_footage` it onto a shot first → `quote_repaint_shot_video` → `repaint_shot_video` (prompt + reference image; with no `reference_image_url` the shot's character portrait is used — without any reference the person is redrawn too) **or** `quote_reskin_shot_video` → `reskin_shot_video` (`first_frame_url` = an already-reskinned first frame that matches the original first frame's composition, pose and props exactly) → `get_repaint_status` with the returned id until `completed` → `restore_repaint_original` to switch back (free). Motion, lip timing and camera move follow the source and its audio is carried over. ≤5 s is one charge (billed as 5 s); with long-shot relay enabled, 5–~14 s is split into ≤3 relayed segments, one charge each, and the quote is the total and the charge cap. A quote is tied to the shot's current video — if that video changes, generating returns 409; re-quote (free) | `edit_video_shot` to swap the whole cast or wardrobe (it edits by instruction); clips longer than ~14 s (`split_shot` first) |
+| `review_storyboards` reports `forbidden_term` (light / color-temperature / BGM wording in the shot text) | every hit is listed at once in the finding's `hits` (`shot`, `field`, `term`) → `get_shot_prompts` to read that field (`description` is returned there too) → `update_shot` to remove the term from that field; if the intent is lighting, move it to `update_shot`'s `lighting` field (not scanned, and video / image generation read it) → `review_storyboards` again | rewriting the whole shot blind, or putting the lighting back into `description` / prompts in other words |
 | A scene plate that came out wrong (backdrop, era, light, layout) | `get_scene_prompt` → `update_scene` (`image_prompt`) → `quote_regenerate_scene_image` (tell the customer the points — it includes the lighting reference the platform auto-derives after a plate change when no wide-shot frame exists) → `regenerate_scene_image` with that `quote_id` (after `upload_scene_image` the platform does **not** auto-create the lighting reference — if the reply says `lighting_ref.needed`, quote it with `quote_scene_lighting_ref` and, once the customer agrees, `generate_scene_lighting_ref`); already-rendered shot frames still anchor on the old plate, so regenerate those shots too | re-running `generate_scene_images` (it only fills scenes that have **no** plate — it will not touch this one) |
 | A voice sample / a required voice | `clone_voice` → `speak_with_voice` → `set_character_voice` / `assign_voices` | cloning without the rights-holder's consent |
 | No sample, but a character already speaks in a rendered shot | `set_voice_anchor_from_shot` (free; native-audio dramas: locks the voice for later videos) → regenerate `stale_shots` | picking the voice after the whole episode is rendered |
@@ -1617,7 +1618,11 @@ to close") tells the vendor to fit that entire sequence into each 3-second shot.
 - **Read back (all free)**: `list_dramas`, `get_drama`, `get_characters`,
   `get_scenes`, `get_assets`, `get_jobs`, `get_pipeline_status`,
   `get_autofill_status` (progress of the background completion batch started by an
-  import's `auto_complete` or by `autofill_storyboards`), `get_cost_estimate`,
+  import's `auto_complete` or by `autofill_storyboards`, plus a counts-only receipt:
+  `skipped_no_gap`, `fields_written_counts`, `contract_coverage` — completion writes
+  internal fields that `get_storyboards` does not return, so report these counts),
+  `quote_autofill` (quote `autofill_storyboards` before running it; tell the customer
+  `estimated_points` / `typical_points` and run only after they agree), `get_cost_estimate`,
   `get_budget_status`
 - **Cut QA (all free)**: `scan_intra_shot_cuts` (vendor cutting inside a shot —
   run this first on any "切太快" report), `recommend_trim_window` (action ends up

package/dist/i18n/tools.en.json

@@ -14,7 +14,7 @@
   "get_bulk_import_spec": "Get the **bulk-import JSON format contract**: full markdown template + template(placeholder template; every <...> must be replaced) + filled_example(finished reference; copy the structure directly) + external_prompt(task prompt to forward to an external AI/your own tool) + checklist + enums(valid values for mode/role/char_type/shot_type/action_motion_class; same source as validator) + limits(array and field caps). Free · static · no credits. ★When to use: before **bulk_import_storyboards**, when the client can export structured data from their own tool/spreadsheet, or wants an external AI to output JSON that creates this episode’s storyboards in one pass. ★Difference from import_storyboard_table(one text line per shot): use JSON if the data is already structured; use that tool if the client has a text storyboard table. ★Contract basics: top level { mode, episode_meta?, characters?[], scenes?[], storyboards[] }; characters[]/scenes[] are **reference declarations only, not persisted**(characters/scenes/props are extracted from the script by extract_assets; import binds by name to existing assets, so extract first, then import; props are not characters—do not put them in characters[]); storyboards dedupe by storyboard_number; bound_characters references characters[].name(or existing character name in this drama), bound_props references prop names in this drama’s prop library(if omitted, prop names appearing in body text are auto-bound), scene_ref references scenes[] location+time—unresolved references are **silently skipped**; each shot needs at least one of action/description; dialogue is `speaker: line`, no stage directions(they will be spoken); image_prompt/video_prompt may be empty(platform builds a base description from shot size + scene + lighting + action); if supplied, they are used **verbatim** and never overwritten. frame_visual_contract is an internal frame contract and ignored on import.",
   "check_bulk_import": "**Pre-import JSON self-check**: validates with the **same zod schema** as the import endpoint, then converts the import route’s \"silent behavior\" into warnings. Returns errors / warnings / stats(shot count · total seconds · characters · scenes · mode · promptsToBuild). Free · not persisted · repeatable. ★errors = import would 400 or overwrite incorrectly: invalid JSON / schema failure(with per-item path, e.g. storyboards[2].shot_type) / duplicate shot numbers + leftover <...> placeholders; must be zero. ★warnings = import continues but something will be missing or have side effects: unresolved bound_characters/scene_ref(silently skipped) / no action+description(dead shot) / dialogue contains camera language(will be spoken) / single shot >15 seconds / action_overload(too many independent actions for the duration—about 2 beats in 3 seconds, 3 beats in 5 seconds; one main action per shot; split new events into next shot; can ignore if multiple people each do one thing) / shot-number gaps / mode=replace(replaces all storyboards in this episode) / no character binding at all(characters will drift). ★If this is green, bulk_import_storyboards should pass in one call.",
   "bulk_import_storyboards": "Create/update this episode’s storyboards from one structured JSON payload in a single pass(dedupe/update by storyboard_number). **Does not create characters/scenes/props**—they are extracted from the script by extract_assets; import only binds by name to existing ones(receipt characters_unmatched / char_refs_unresolved lists unbound refs; extract first, then import). Free · deterministic · no AI call. ★Three steps: ① get_bulk_import_spec for contract and finished example; ② check_bulk_import until errors=0; ③ call this tool. ★mode=merge(default) keeps shots not mentioned; mode=replace replaces **all** storyboards in this episode(old shots and generated images/videos are archived and recoverable, not auto-relinked)—replace requires explicit client consent in any execution mode, including autopilot. ★If you want to write image_prompt / video_prompt yourself, put them on each shot; they are used **verbatim and not overwritten**. Empty shots get platform-built base descriptions(receipt base_prompts_built says how many); after import use get_shot_prompts to read and update_shot to edit. (frame_visual_contract is an internal frame contract and ignored on import.) ★After import: speech timing may raise durations for shots whose dialogue does not fit(receipt speech_duration_raised / speech_duration_overflow). ★After import, **auto-completion is on by default**(auto_complete, follows the drama’s \"auto-complete auxiliary assets\" switch, default on): one background batch fills empty professional fields across all shots(AI, fill-empty only) and expands full image/video prompts only for shots with no image_prompt that used a platform base description(client-written prompts are never overwritten; receipt auto_complete.prompt_shots). Text step is postpaid by token; **tell the client before calling**. If you only want import, pass auto_complete:false(**cost: no composition contract is generated**—visible headcount, subject type character/part/prop, and shot size can only be inferred from text; hand insert shots and off-screen narration shots are easily rendered/audited as full characters; receipt contract_coverage lists missing shots **and gives current viable fixes; trust its note**—when contract auto-complete is enabled it calls autofill_storyboards). If receipt started=true, poll get_autofill_status until done **before** review_storyboards.",
-  "get_autofill_status": "Check this episode’s background completion batch(import auto_complete / autofill_storyboards AI one-click fill-empty): { status: running|done|failed|idle, total, done, ok, fail }. Free. ★After import with auto-complete, wait for status=done before review_storyboards / generate_frames—the batch rewrites shot fields one by one, so any earlier review_token becomes invalid when content changes. Each shot takes tens of seconds; a full episode can take 10+ minutes. Poll interval ≥10 seconds; do not tight-loop. If status=failed or fail>0, inspect fail count; a few failed shots can be rerun with autofill_storyboards(fill-empty only; already-filled fields are not charged again).",
+  "get_autofill_status": "Check progress and receipt of this episode’s background completion batch (import auto_complete / autofill_storyboards AI one-click fill-empty): { status: running|done|failed|idle, total, done, ok, fail, skipped_no_gap, fields_written_counts, contract_coverage }. Free. ★The receipt is the evidence of what completion wrote: it writes internal narrative fields and composition contracts that get_storyboards / get_shot_prompts **do not return**, so seeing no change on read-back is normal—use these counts instead. skipped_no_gap = shots with no gaps that were skipped (no AI call, no charge); fields_written_counts = number of shots written per field name (e.g. {shot_intent: 44}); contract_coverage = {total, missing_contract} = image-generating shots in this episode and how many still lack a composition contract. Counts only, never field contents—tell the customer these numbers as-is. ★After import with auto-complete, wait for status=done before review_storyboards / generate_frames—the batch rewrites shot fields one by one, so any earlier review_token becomes invalid when content changes. Each shot takes tens of seconds; a full episode can take 10+ minutes. Poll interval ≥10 seconds; do not tight-loop. If status=failed or fail>0, inspect fail count; a few failed shots can be rerun with autofill_storyboards (in fill-empty mode, shots already filled are skipped with no AI call and no charge; only unfilled shots are billed).",
   "adopt_external_script": "[Exit A] Adopt a draft produced by an external AI from the template **directly as the shootable script**; StarReel AI does not intervene. Seconds, no charge. ★Hard pre-gate: server first runs a format self-check; non-empty errors return 400 with the list unchanged—use check_script_format until fully green before this tool. warnings do not block(same strictness as default lenient save_script). ★When to use: client demands \"do not change a single word\" and the external AI already produced a complete draft with production-layer annotations. ★When **not** to use: client only organized the story layer(no [角色档案]/[外貌])—direct adoption would leave downstream without character profiles and appearance anchors; use Exit B(set_script + rewrite_script) so the platform adds annotations. ★Difference from edit_rewritten_script: that is a point-edit channel for an **existing rewritten script**(400 if rewrite was never run). This tool is the **first script ingestion** channel, only for \"the script was prepared externally\". It is not a backdoor around rewriting—the server gate decides; if it fails, it fails. Old script is saved into previous_script; if adopted by mistake, recover with get_script(include_previous=1).",
   "extract_assets": "Extract characters/scenes/props from the shootable script (writes three tables in one pass; the foundation for downstream consistency). Backend async; text step postpaid, no arrears. Prerequisite: rewrite_script has produced the rewritten script (mandatory for new projects; character profiles must be extracted from the rewritten script to stay self-consistent with script and storyboards). ★Minute-level backend task; use get_run_status to decide whether it is still running; do not treat 60 seconds as failure.\n★【Rewritten-script review hard gate · free】Before this step, you must call review_script first: report every returned finding to the customer verbatim (each with code=issue type, shots=hit shot numbers, action=which tool to use to fix), then pass review_token into this tool. Skipping review will be rejected with 400. If content changed after review → token auto-invalidates; review again once (still free). If there is error, block by default; only if the customer is informed and insists on continuing as-is may you pass acknowledge_review:true—pushing forward with known issues will likely produce unusable footage and is still charged normally (acknowledge_review only means the customer personally agreed; every execution mode still requires explicit customer consent, including managed mode); in managed steps, only allowlisted heuristic error can pass with auto_acknowledge:true (platform decides; if not allowlisted returns 400 AUTOPILOT_MUST_ASK:must_ask_customer=true then pause and ask customer, false = self-fix per action then review again). ★Gray-release dramas also require you (agent) to QC personally: inspect assets item by item per agent_qc returned by review_script, answer the checklist, and submit via submit_qc_report before this token can be used (not submitted → 400 agent_qc_required; bad report → blocked by default like error).★★【Execution-mode master rules · read before starting】For dramas with a confirmed opening plan, first get_production_plan and read autopilot: pro mode (pro, or autopilot.enabled=false, or no confirmed plan) → every later 'tell customer, proceed only after customer agrees' is literal; one-click (auto) / hybrid (custom) managed steps → do not ask customer: follow get_pipeline_status next_action in order, self-confirm quotes within the hard budget cap, after each layer still run review_* + agent QC, and let review/QC autopilot.next_step tell you how to proceed at this layer (allowlisted error with auto_acknowledge:true / which self-fix round / pause to ask customer); for choices follow autopilot.default_decisions; autopilot.never_auto_actions(mode=replace, deleting assets, overwriting customer-uploaded images, allow_era_ref_conflict, cloning voices, delivery acceptance) always require customer consent in every mode. ★In managed steps, regenerating frames for the same shot (generate_shot_frame) max 3 rounds, regenerating video (regenerate_shot_video / edit_video_shot) max 3 times; at the limit the platform returns 409 AUTOPILOT_SHOT_REGEN_LIMIT (not charged)—do not bypass with another tool; stop and ask customer: if customer approves more, quote normally and pass customer_approved_more_regens:true when generating (grants another limit); if customer accepts current state, continue, and pass acknowledge_review:true on the next paid call to record confirmation. 402 BUDGET_WOULD_EXCEED / 409 MODEL_UNAVAILABLE / 409 AUTOPILOT_SHOT_REGEN_LIMIT / 400 AUTOPILOT_MUST_ASK with must_ask_customer=true / non-empty pending_questions → pause and ask the customer the pending_questions from get_autopilot_log **summarized all at once**; customer-controlled steps follow pro mode. After final output, deliver to customer using the get_autopilot_log report. ★Three-tier execution strategy (do not mix the three when asking customer):①【Basic project settings · free · mandatory foundation · set when creating drama, do not create an empty shell】project_type/setting_brief(worldview · ERA LOCK)/ethnicity/aspect ratio resolution, plus consistency anchors cinematography_prompt(photography DNA)·art_bible(art bible)·visual_lock(visual lock)—all free, foundation for full-pipeline consistency; if not set properly, all later generation drifts and rework costs again. Set directly via create_drama/update_project_settings. ★visual_lock/art_bible only write image-level/world-level locks (shot language · environment · art tone · forbidden elements), **never pin clothing/hairstyle/appearance details for specific characters**—the single source of truth for character appearance is the character profiles produced by extract_assets (change via update_character); writing both places inevitably conflicts, character sheet vs profile, concept image vs visual lock, consistency gate rejects by character sheet → concept images/shot frames get **structurally rejected in chains**, no amount of rerolls passes, pure waste.★★【Era contract · free · required before image generation for non-contemporary or surreal subjects】The worldview in setting_brief is **free text**; it can provide background but cannot constrain images. The real override over visual_lock is the **structured era contract**: use set_era_contract to set this episode’s tier (modern/republican/historical/fantasy/surreal), get_era_contract to read current value. **Unset means empty**—the platform can only rely on prompt text to fight drift; modern objects entering ancient images or scene images drawn in the wrong era is usually caused by this being empty. ★Xianxia/fantasy/immortal-realm content **must use fantasy, not historical**: historical requires textual realistic accuracy and will classify glowing aura, floating landforms, and non-human transformations as 'era-inconsistent', pulling images toward realistic ancient style. To make an exception for one scene (e.g. a modern flashback inside a fantasy world), use update_scene era_contract to override at scene level; scene-level beats episode-level.★★【Per-stage review protocol · all free · main line to prevent unusable film, not optional】After every stage output, review first, report conclusions verbatim to customer, then proceed. **Five hard gates (not passing causes 400; ④⑤ and agent QC below)**:① after rewritten script → review_script (before extract_assets / generate_storyboards);② after storyboards → review_storyboards (before generate_frames);③ after shot images → review_frames (before generate_videos). Each review returns review_token; pass it with downstream paid tools; explain findings one by one to customer (code=issue type · shots=hit shot numbers · action=which tool to fix), fix per action, then **review again** before proceeding. If content changed after review → token auto-invalidates; review once again (free). If error exists, block by default; only if customer explicitly understands and insists may pass acknowledge_review:true (acknowledge_review only means customer personally agreed; every execution mode still needs explicit customer consent, including managed mode); in managed steps, only allowlisted heuristic error can pass auto_acknowledge:true (see 【Execution-mode master rules】). **Soft guidance (non-blocking but strongly recommended, also free)**: before image/video generation run run_precheck (catches shots vendors will definitely reject, preventing wasted credits)—★**to run whole-drama first use run_drama_precheck** (drama-level summary · free · read-only): get issue counts per episode and attention (which episodes to fix first), prioritize by it, then run run_precheck on those episodes for details; do not blindly scan episode by episode. Note not_covered—drama-level skips the only lint that requires images, so '0 whole-drama issues' ≠ 'everything checked'. ★After findings, do not blindly edit yourself: plan_precheck_fix lets platform compute proposals → apply_precheck_fix writes them (choose which to adopt; managed=adopt all per execution mode: pro mode / no confirmed plan / this step customer-controlled → tell customer and proceed only after consent; do not decide for customer; managed step → do not ask, follow get_production_plan autopilot.default_decisions, platform logs in get_autopilot_log); after storyboards run get_health_report; after character sheets use get_characters to verify every appearing character has image/sheet; after frames use get_storyboards to inspect frame_status and fail_reason/fail_hint (fix failed shots before continuing; do not make video from bad frames); after videos inspect video_status similarly; before final cut use get_pipeline_status to confirm no missing shots. ★**Never generate straight through**: pushing layer after layer without review amplifies issues at every layer, and the whole episode may need redoing—every repeated image/video generation is real charged. Reviews are all free; catching issues costs nothing. ★Clothing changes are structured bindings: update_shot active_wardrobe_id (whole shot) / appearance_wardrobes (multi-character per role), can override or clear—editing text descriptions does not change clothes, and do not send customer to web UI.②【Production mainline · sequential, no skipping · ★storyboards before assets】set_script→rewrite_script→★review_script→extract_assets→storyboards(first storyboards · pure-text shot breakdown)→★review_storyboards→★Script discipline (endpoint-enforced, cannot bypass): all raw material (synopsis/outline/finished script all count) goes into set_script, **must go through rewrite_script to produce AI rewritten script**—pasting your own finished script directly into edit_rewritten_script to bypass rewrite will be rejected 400 (no rewritten script means nothing to edit); extract_assets likewise requires the rewritten script. All post-rewrite changes must use the structured format of AI output: edit script via edit_rewritten_script (polish/correct), character profiles via update_character, storyboards via update_shot/replace_shot_dialogue—do not go back and replace the whole script or write another truth source in settings; dueling truth sources are the top cause of consistency failures.★★**After rewrite_script succeeds once, do not rerun rewrite_script**: it restarts from the original full manuscript, discarding all current edits, and the new version is not guaranteed to preserve prior fixes (three-version tests oscillate). To fix, use edit_rewritten_script point edits (get_script full text → change only those scenes, copy the rest verbatim → submit whole text), free, seconds-level, deterministic; if rerun by mistake use get_script(include_previous=1) to recover previous version. generate_portraits_and_sheets(character sheet + concept images · build after storyboards: first use delete_character to delete appearing characters who **do not appear in any episode of the whole drama and are not planned for later episodes** before image generation—characters are drama-level; once deleted, no episode’s frame assembly includes them; this cannot be judged from current episode storyboards alone, keep if unsure (one character sheet costs far less than losing identity anchors in a later episode); the character sheet step generates images for **all undeleted appearing characters**, ignoring storyboard bindings, so kept characters are charged normally; voiceover characters (char_type=voiceover narrators/phone voices) are not image-generated and have no storyboard binding; **do not delete** them, dubbing relies on them)→★generate_color_script(color script · unified grading)→assign_voices(assign voice types; if narrator/character library has no fit → design_voice by description, then set_character_voice after sample approval)→★quote_scene_images+generate_scene_images(empty scene plates · required before frames)→frames(default first frames only)→★tail_frame_plan(free · which shots need independent last frames)→frames(frame_type=last_frame)→★chain_frames(seams · dry_run:true first)→★review_frames→videos→generate_tts→compose;★★【Do not skip seams · after last-frame completion, before review_frames】After first/last frames are ready, call chain_frames(dry_run:true, free) to view platform seam plan and estimated_points (copy free, reframe one image per seam), tell customer and execute without dry_run after confirmation; already chained seams are not planned, repeated calls do not charge again. ★After execution, reframe is **async image generation**: use get_jobs to watch first_frame generation rows for planned shots; wait until all pending/processing finish (failed: check error_msg; use chain_shot_from to rechains a single seam), then review_frames—while old first frame remains, get_storyboards frame_status is still ready, so it cannot show whether the new image is generating or failed. After get_jobs is clear, use get_storyboards to verify seam_state: replacement images rejected from promotion will not remain in get_jobs; that seam remains unchained (review_frames also reports seam_unchained); rechains per its action. If you review before landing, you reviewed old first frames; the token becomes invalid after new first frames land. Episodes generated shot-by-shot/batch likely were not auto-chained (get_storyboards seam_state=unchained means so); skipping does not error—the cost is mismatch between adjacent shots and jump cuts throughout final cut.★★【Do not skip last frames · frames are two passes】generate_frames defaults to first frames only. About 30% of shots have **end state ≠ start state** (large camera moves/object released/state changes) and need an independent last frame; the criterion is platform-side and cannot be guessed from storyboard text—therefore after first frames finish, you must call the free tail_frame_plan once to get per-shot list, then fill frame_type=last_frame (shots_needing_last_frame in generate_frames response is this count; if not 0 do not go straight to review_frames). Skipping does not error or block—the cost is that videos for those shots have only one anchor, so the model invents the end state: actions fail, large camera moves end back at the starting composition. Production tests: in 32 episodes, 30 generated only first frames for the whole episode, and 23 went all the way through video.★★【When last frame is blocked by TERMINAL_DESC_GATE: edit text, do not retry】For shots reporting 'this shot is marked as state-changing, but nowhere describes what it looks like at the end', use update_shot to write **what the image looks like at the end** into that shot’s last_frame_prompt, then regenerate last frame. ★Better: **avoid hitting it**—before last-frame generation, first inspect get_storyboards; shots with terminal_desc_missing=true are listed for the whole episode at once (free), fill their last_frame_prompt before generating and avoid any rejection. Per-shot query also exists (get_shot_prompts same field), but do not do that for dozens of shots. Retrying without filling the description is useless—the vendor prompt has no terminal-state section, so it still draws the first-frame pose and will be rejected again by terminal review, and each attempt is charged normally (production tests: last-frame success 9.4% without terminal description, 40.2% with it). Only use generate_shot_frame allow_missing_terminal for shots that truly should be 'almost unchanged'.★★【Do not skip empty scene plates · this step is often missed by third parties】Scene images are **background anchors**: every shot frame in the same scene anchors on it. Frames can still generate without plates, with no error and no block—the cost is **the first shot of every scene has no background anchor at all** (platform fallback is 'if missing image is discovered, generate one in background'; that plate is used by subsequent shots in the same scene, but the triggering first shot cannot wait for it), and the first shot is often the tone-setter; later backgrounds drift too. It is paired with character sheets: character sheets anchor people, scene images anchor locations; whichever is missing drifts. Progress self-check: get_pipeline_status generate_scene_images step (completed/total); review_storyboards also reports scene_plate_missing before frame generation.★★【Do not skip concept images · character sheet ≠ concept image】A character sheet is a single-person identity anchor; concept image is the platform-generated multi-angle turnaround from the character sheet, the root consistency anchor for shot frames/videos (angles, lighting, clothing rely on it). Frames can still generate with only character sheets, with no error—the cost is characters drift when angle changes and clothing differs every shot. Use generate_portraits_and_sheets for one-click two-step (each call advances one step; after character sheets are complete, call again to create concept images), or generate_character_sheets after character sheets. Progress self-check: get_pipeline_status generate_character_sheets step (completed/total, portrait_only=characters with character sheet only). ★Customer images: character sheet must be single person, single image, single angle, no text; scene image must be empty location with no people and no overlaid text—set_character_portrait/upload_scene_image auto-checks and rejects invalid with 422. ★Do not create character visuals/prop concept images/action templates before storyboards—storyboards are pure text and do not depend on any images; after storyboards the cast is known, delete non-appearing characters before image generation to save cost (action templates must be after storyboards anyway). Paid steps follow existing quote confirmation flow. Ads additionally need add_product+generate_product_sheet; MV uses set_mv_lyrics→generate_mv_story→generate_mv_script (if no finished song: after set_mv_lyrics first suggest_mv_song_styles→quote_mv_song→generate_mv_song→get_mv_song→select_mv_song).★★【④⑤ two new hard gates + agent personal QC】④Asset layer: after character sheets/concept images/scene images/prop images finish → review_assets (before generate_frames / generate_shot_frame; its token is asset_review_token and passed together with storyboard-layer review_token);⑤Video and dubbing layer: after video/dubbing/music finish → review_videos (before compose_episode); after final cut, run review_final_cut. ★After every review layer you must also **personally QC**: inspect assets item by item per returned agent_qc (download images, sample video frames, read text), answer checklist item by item, submit via submit_qc_report; for gray-release dramas, a review_token cannot enter the next step without the report; bad report is blocked by default like error. Platform judges existence; you judge likeness and correctness.★★【Single-shot path is also gated by reviews】: generate_shot_frame also requires storyboard-layer review_token; regenerate_shot_video also requires shot-image-layer review_token—do not expect single-shot tools to bypass review (production observed a workflow generating all 25 shots with single-shot frame tools and no gates; all final rework was fully charged). Generated frames **do not** invalidate the storyboard-layer token (fingerprint only covers storyboard content); only storyboard content edits—update_shot / import / add-delete shots / change character bindings—require one more review (free, seconds). Generating one shot’s frames invalidates the **shot-image-layer** token; before regenerating the next shot video, review_frames once.★★【For native-audio dramas, lock voices before batch video】When use_clip_audio=true(default), each shot’s voice is sampled independently by the video vendor; if not locked it will drift, and unifying after the fact requires per-shot regeneration. If customer-authorized voice source exists → clone_voice+set_character_voice before generate_videos; if not → for each speaking character first use single-shot quote_regenerate_shot_video→regenerate_shot_video on a shot with longer dialogue (generate_videos has no per-shot parameters and only batch-fills missing-video shots; single-shot regeneration also works for shots without video), then use set_voice_anchor_from_shot to select the voice from that shot, then generate remaining shots. get_pipeline_status native_voice_anchor lists unanchored_speakers not yet locked and stale_shots generated before locking that need regeneration. ★Worldview concept image = required by default (improves whole-drama consistency; many third-party platforms miss it): after storyboards, call generate_world_concept by default; still follows quote confirmation (tell customer estimated credits, confirm before charging)—do not charge silently and do not skip.★★【Color script】generate_color_script unifies grading across the whole film; run after storyboards and before frames; text step has no quote_*, postpaid by usage. Progress self-check: get_pipeline_status generate_color_script; review_storyboards reports color-script gaps. ★Action templates are optional experimental: call generate_motion_templates(action template extraction) after storyboards only if customer explicitly requests; text step postpaid by usage. First/last frames do not inject action templates by default; missing templates need not be made up and do not block image/video/final cut. Video templates may be cropped; delivery or quality gain is not guaranteed. ★Scene Bible(detailed per-scene setting) comes **after scene image generation**—refine based on generated scene images (no MCP tool yet; do on official site); do not make Scene Bible before scene images. ★Audio defaults to video native audio (use_clip_audio default on; skip TTS and use AI video’s built-in audio): when creating/updating drama settings, AI should proactively tell customer 'default uses video native audio; set use_clip_audio false if TTS dubbing is needed' and let customer choose. ★Default image model is ChatGPT Image 2.5 Flare(gpt-image-2.5-flare · unified whole-drama style): set image_model in create_drama/update_project_settings; omitted means default. Options: gpt-image-2.5-sunburst(same price · better Chinese glyphs/details)/gemini-3.1-flash-image(Banana 2 · 71 credits)/gemini-3-pro-image(Banana Pro · finer · 175 credits)/gemini-3.1-flash-lite-image(Banana 2 Lite · cheap · 31 credits)/doubao-seedream-5-0-260128(Seedream5.0); generate_frames can temporarily override one run. ★2.5 charges by input volume (base 11 credits + 18 credits per reference image); Banana/Seedream are flat price—shots with many reference images must include this in budget. ★Four video engine choices (drama-level · when AI creates a drama it must proactively guide engine choice by drama type and quote price differences for customer decision):【Decision tree】★Call get_production_menu first: whether each engine is currently available for realistic live-action / anime and other subject classes is decided by the availability it returns (unavailable entries carry customer_note and expected_until—relay them to the customer verbatim); do not assume from this description that any model is unavailable, and do not recommend models the menu marks unavailable. ①Realistic live-action drama→seedance-2.5 (strongest instruction following / face detail · pricing baseline 720p = 1×), or hailuo-3 when budget-sensitive (about 1/3 cost, 720p ≈0.33× · strong fidelity editing · can output 2K · but about 6 minutes per shot); ★After choosing, pass video_engine explicitly in create_drama / update_project_settings (the one the menu lists as available for this subject class)—if omitted, the platform only applies a style-based fallback stamp that is not guaranteed to be an engine the menu marks available; ②Stylized/animation/3D cartoon dramas · empty shots · product shots→wan3.0(about 40% price, 720p ≈0.4× · max 30 sec · minimum 2 sec billing · about 2 minutes per shot), use wan3.0-prime for rush delivery (720p ≈0.6× · about 1 minute); ★Always quote prices as relative multiples (seedance-2.5 720p = 1×); call quote_* before giving the customer absolute credits;③★Never choose wan3.0/prime for realistic live-action dramas—WAN output-side human-face review consistently rejects at 720p+, retry cannot fix;④★★For narrative dramas (dialogue, continuous story, stable shot rhythm), choose wan3.0/prime cautiously: WAN may **hard-cut camera positions inside a single storyboard shot** (tests: 11/12 shots had intra-shot jump cuts, versus seedance-2.5 1/6 and hailuo-3 0/5); final cut feels like 'the image jumps around within one shot, cuts too fast'. This is weak vendor instruction following and cannot be blocked at prompt level (our negative constraints are already included and tested ineffective); afterwards the only fix is regenerating with another engine. WAN fits stylized/empty/product shots that are inherently short and fast; for stable single-shot narrative choose seedance-2.5 or hailuo-3. After generation use scan_intra_shot_cuts to check;④b★★Dialogue-heavy dramas should choose hailuo-3 cautiously (different from above 'intra-shot self-cut'; this is about **whether all lines are spoken**): native-audio engine stops speaking at shot end and may ad-lib; tests show 'dialogue not fully spoken' rates hailuo-3 **50%** (26 shots, all generated after 8-30 native-audio repair, so engine itself), seedance-2.5 **23%** (294 shots); wan3.0 sample **insufficient, untested** for this dimension (26 samples all before same repair, only 2 after)—do not infer WAN is bad. When customer reports 'line not finished', first run scan_dialogue_coverage by family; do not default to lengthening shots (even among sufficiently long shots, 32% still missed full lines);⑤★Shot length control (all engines, especially WAN): keep each shot 3~5 seconds. The longer the shot, the more room the model has to improvise—test: one 16-second shot (neighbors all 3~5 sec) changed scenes 4 times within the shot; character disappeared for 4 sec then grew back from frame edge; customer sees 'a person appeared from nowhere'. For long performances, split into multiple short shots and connect with frame chains; do not write single shots over 15 sec;【Resolution decision】Draft/iteration: WAN dramas 480p (≈0.2× cheapest), others 720p; final delivery: seedance dramas 720p (for HD delivery choose 1080p, about 2.5× 720p), hailuo-3 dramas 1080p(=2K · ≈0.53×), WAN dramas 1080p(≈0.8×); hailuo-3 has no separate 480p tier (if selected, billed as 768P); set video_engine/video_resolution in create_drama/update_project_settings, ★must be decided before video generation—switching does not retroactively affect generated shots, and mixing within one drama causes style/identity drift. ★Slow image generation ≠ failure: each image takes tens of seconds to minutes; whole episode may take 10+ minutes. Poll get_storyboards frame_status—pending=still generating (wait patiently; do not call generate_frames again and waste credits), ready=done, failed=real failure. ★To change one shot’s image / make new character sheet take effect, use **single-shot regeneration generate_shot_frame** (platform automatically includes identity anchors, scene/prop references, style anchors for that shot to preserve whole-film consistency); generate_frames only batch-fills shots with **missing frames** and skips shots that already have first frames (normal, not 'rejected'). ★After first frames finish, call free tail_frame_plan to see which shots need independent last frames, then batch-fill with frame_type=last_frame. After changing a character sheet (set_character_portrait), stale_frames in the response are shots polluted by the old image and must be regenerated shot by shot. ★Never use external tools to make first/last frames and upload_shot_frame to 'change the image'—external images lack identity/style anchors, so characters/clothing/style will drift; that is the real source of unusable film. upload_shot_frame is only for customer-owned real footage.③【Optional enhancements · AI proactively informs customer · do only after quote confirmation】Art bible generation / visual lock extraction / scene groups / lip sync / posters / sound effects / music / subtitle translation—these improve consistency/quality and are mostly paid. ★AI should proactively tell customer these are available and provide quote (whether to run depends on execution mode: pro mode / no confirmed plan / this step customer-controlled → tell customer and proceed only after consent; do not decide for customer; managed step → do not ask, follow get_production_plan autopilot.default_decisions, platform logs in get_autopilot_log); neither silently skip nor charge without authorization. ★Two locking rules:①**Aspect ratio** is set at create_drama, locked at drama level; all later image/video/final cut use it, **do not change midstream** (generated content will have inconsistent aspect ratio and drift); default 9:16 if unset.②**5-7 sec per shot in shot breakdown is normal length optimized for AI video**, do not wrongly re-split because 'shots are long'—generate_storyboards **replaces all storyboards in the episode**, wasting generated images; backend blocks when storyboards already exist unless confirm_replace is supplied. ★Rewrite fidelity (default auto smart routing): when the set_script original is **already script-shaped** (scene headers/dialogue-line structure), rewrite_script automatically uses two-pass fidelity—customer dialogue is machine-gate locked line by line (missing one line causes internal rejection and redo), AI never adds plot; dramaturgical gaps (hooks/emotional anchors) are not auto-filled, but written into dramaturgy_suggestions returned by get_script (whether to adopt follows execution mode: pro mode / no confirmed plan / this step customer-controlled → tell customer and proceed only after consent; do not decide for customer; managed step → do not ask, follow get_production_plan autopilot.default_decisions, platform logs in get_autopilot_log). If original is novel/outline, it auto uses creative rewrite (AI lays hooks/emotional beats); both customer types get intended behavior, no manual switch needed. To override default use update_project_settings rewrite_pipeline(auto/two_pass/single_forced) and fidelity_enforce(1=hard fidelity gate). When customer says 'AI changed my script too much':①confirm full original entered set_script;②set rewrite_pipeline two_pass to force fidelity, then rerun rewrite_script;③after customer confirms character appearance, use update_character profile_locked=1 to lock profiles and prevent later extraction from overwriting appearance and changing faces in character sheets. ★If customer wants to rewrite script on **another AI platform** (common: third-party model says our rewrite 'changed too much', customer wants control): first call get_script_format_spec to obtain platform-accepted format contract (markdown template + external_prompt directly forwardable to external model + blank skeleton), give external_prompt+template+customer original to that platform; after receiving organized draft, **first call check_script_format** self-check (free · rule-only · no model); after errors are zero, two exits:【A】adopt_external_script directly adopts as shootable script (our AI not involved · seconds · no charge, requires external draft contains production-layer annotations);【B】set_script puts it back into original slot + rewrite_script two-pass fidelity (choose when external only handles story layer; platform fills annotations; customer-written annotations are stripped and rewritten on this path). ★Do not put external organized draft into edit_rewritten_script (400 if rewrite not run), and do not skip check_script_format before ingestion—noncompliant scripts are still blocked by gates, wasting a run.★★If the customer’s material is **already a finished storyboard table** (per-shot durations/framing/camera moves), neither path applies—use import_storyboard_table directly to create storyboards and skip rewrite and shot breakdown. The rewrite path will strip durations/framing/camera moves/STYLE/text cards as non-story content (production test: 8 shots 36 sec → 20 shots 109 sec). ★Both import channels are deterministic—wrong input is created as-is, so **get contract, self-check, then import**: storyboard table path uses get_storyboard_table_spec → check_storyboard_table → import_storyboard_table; if customer tools/spreadsheets can export structured data, or external AI can directly produce JSON, use get_bulk_import_spec → check_bulk_import → bulk_import_storyboards (only creates storyboards; characters/scenes/props are extracted from script by extract_assets; import binds by name to existing ones—extract first, then import). Both imports default with auto_complete(background AI fills professional fields + image/video prompts, text step postpaid; tell customer before call): if receipt started=true, poll get_autofill_status until done, then review_storyboards—autofill changes shots, so any earlier review token invalidates. Use get_pipeline_status for progress (returns project-type-specific steps).★★【Opening plan · first thing after creating drama · free】First ask basic video parameters (project type/aspect ratio/episodes × minutes per episode/subject: realistic_human realistic live-action · stylized anime/stylized · product/empty shots · native audio or dubbing), then call get_production_menu to get **currently available** three packages (effect description + real-time estimate including text steps + recommended budget + excluded items) and explain to customer—do not recommend models from memory; some models may be temporarily unavailable for some subjects, menu already filters them. Then let customer choose execution mode (pro self-control every step / auto one-click final cut / custom managed selected steps), budget budget_points, and automatic overrun cap (0/10/20/30/50%): **default recommendation** budget=selected package recommended_budget (upper bound excluding text ×1.2 plus text estimate), overrun=menu recommended_overrun_pct(30%)—QC regenerations are not in upper bound and are covered by overrun; recommendation is only recommendation, final numbers are customer’s, and if customer changes them use customer values. Explain summary clearly; only after explicit customer consent call confirm_production_plan(customer_confirmed:true)—never confirm for customer. For dramas with confirmed plan: before paid dispatch the platform rechecks model availability and quote upper bound ≤ budget×(1+overrun%); if blocked (402 BUDGET_WOULD_EXCEED / 409 MODEL_UNAVAILABLE), stop and ask customer; do not change model or budget yourself. confirm returns 403 PRODUCTION_PLAN_NOT_OPEN = this drama not open (gray release), follow original step-by-step confirmation flow.\n★【Post-completion QC】After extraction, use get_characters / get_scenes / get_props to verify profiles (names, appearance, era) match the customer script; after character sheets/concept images/scene images/prop images finish, review_assets handles coverage and quality review uniformly.",
   "get_pipeline_status": "Check the complete workflow progress for one episode(script_rewrite/extraction/storyboards/voice/image/video/compose/music/final merge… each step done/partial/pending/not_required). Follow it in order; do not skip. Free. ★For auto / custom mode(hosted rollout), also returns next_action(action=run do hosted step directly / wait_running background still running, wait for persistence and do not dispatch duplicate / wait_customer client-controlled step or no budget precheck, ask first / ask_customer pending questions, summarize and ask first / done all steps complete—first run tool=review_final_cut, fix issues and re-compose/re-review, then deliver report; tool=which tool to call) and pending_questions; advance exactly by next_action. asset_versions_summary separates generated, input_consistent, review_valid, adopted, and current_effective; do not treat generate_videos.completed as current valid count. ★generate_images / generate_videos status=done only means \"generated\", not deliverable: inspect validity for that step(current_effective / input_changed with changed_by listing drama-level settings, shot itself, or which character changed / input_unknown / unreviewed / review_failed). If validity.note is non-empty, tell the client first. Changed input may not affect visuals; do not regenerate whole episode paid just for that—use get_asset_versions per shot to inspect causes. ★not_required means this mode does not need the step(e.g. TTS steps for native-audio drama, generate_bgm when music disabled), not unfinished; do not fill it. Image/video denominators exclude blocked shots(shots_not_applicable). merge_episode.bgm_stale=true means music is newer than final cut; just compose_episode again.",
@@ -41,13 +41,14 @@
   "apply_precheck_fix": "Persist proposals the customer has **confirmed** (step ②). Free (pure text DB write). Input is the proposal object returned by plan_precheck_fix; just **pass it back unchanged**—only the items the customer checked for acceptance.\n★Before persisting, the platform applies three gates again: optimistic lock (if the DB value changed, skip; never overwrite others' changes), admission recheck, and self-healing gate (if the rule is not green after the edit, skip). So some items entering skipped is normal, not a failure.\n★Returned remaining = how many precheck issues **remain in this episode**, not \"how many fields were changed\". After fixing, rerun run_precheck to confirm, then proceed to image generation.",
   "get_health_report": "(★Recommended · free diagnosis) Read storyboards and output a health report: overlong duration / insufficient motif coverage / problematic shots. Read-only, free. Check before image generation; identify issues, fix first, and avoid discovering them only after the film is produced.",
   "review_script": "【Hard Gate ① · Free】Review the AI rewritten script: dialogue line length, completeness of emotional turns, opening hook, final-scene cliffhanger. ★Before extract_assets and generate_storyboards, this tool must run first—character profiles and storyboards are derived from the rewritten script; issues in the script will be amplified layer by layer into character sheets / image generation / video generation, and discovering them at final cut means redoing the whole episode. Tell the client each item in findings one by one, fix by action (use edit_rewritten_script for script edits), then re-review and proceed with the review_token.\n★【Agent manual QC · Free】The returned agent_qc contains this-layer checklist and the episode materials to inspect (materials, each with a ref; images/videos include links). The platform only judges “exists / present / rejected”; “looks alike / correct / good” must be checked by you: download and open images, sample at least beginning/middle/end frames for videos, read text through with tools. Answer each checklist item (ok / issue / na; for issue, specify which item and what problem), list every inspected ref in inspected, and call submit_qc_report to submit; relay the QC conclusion to the client verbatim. If agent_qc.required=true and no report is submitted, this review_token cannot enter the next step (400 agent_qc_required); a problematic report blocks by default just like a platform error.\n★【Script evaluation and suggestions】materials contains the client original script (script:original) and the rewritten script (script): first evaluate the client’s script itself (mainline, motivation, conflict, pacing, plot holes, filmability), then check the rewritten script; for each issue, provide concrete fixes in submit_qc_report suggestions (if wording can be directly changed, provide before/after). After submission, tell the client the evaluation and suggestions item by item and let the client choose: adopt some / execute all directly / continue without changes (which path to take follows the execution mode: professional mode / unconfirmed plan / this stage controlled by client → tell the client and proceed only after client approval; do not decide for the client; if this stage is autopiloted → do not ask the client, execute according to get_production_plan autopilot.default_decisions, and the platform records it in get_autopilot_log); autopiloted stages by default execute only priority=must items that do not change the client’s original dialogue; all others go into the delivery report. ★[Customer acceptance items] review_frames / review_videos append the customer's registered specific requirements (add_acceptance_criteria) to the checklist (ids start with req:) together with acceptance_criteria: the platform's generic review does not know these requirements, so passing it does not mean they are met. Check each one with your own eyes, answer it in submit_qc_report, tell the customer the result, and let the customer confirm via confirm_acceptance_criterion.",
-  "review_storyboards": "【Hard gate ② · free】Review storyboards: forbidden terms (will be rejected by vendors and waste credits), shot duration distribution, repeated adjacent composition, character blocking drift in the same scene, emotional curve peaks/valleys, key-shot markers. ★Must run this tool before generate_frames—storyboard issues become a whole episode of wasted images once generated and cannot be fixed shot by shot. Fix per findings.action using update_shot/split_shot, then review again. ★A finding with code=forbidden_term includes `terms` (specific matched terms, e.g. \"background music\", \"strong light\") and `shots` (shot numbers); use those two to update_shot and remove them, no need to guess word by word. ★Forbidden terms are judged with **negative semantics**: writing \"no background music\" or \"do not add music\" is not a violation (it is obeying the constraint), so no edit is needed. ★\"Action overload\" in warnings (action_overload, advisory only, non-blocking): one shot contains more independent actions than its duration can support; before video generation, use split_shot to split it or delete actions—slow motion/dissolves cannot rescue it after final cut. Ignore if multiple people each do one thing or it is a naturally continuous action chain.\n★【Agent hands-on QC · free】Returned agent_qc contains this layer's checklist and materials to inspect for this episode (materials, each with a ref; images/videos include links). The platform only checks \"exists / present / rejected\"; \"looks like / is correct / is good\" must be judged by you: download and open images, sample at least beginning/middle/end frames for videos, read text fully with tools. Answer checklist items one by one (ok / issue / na; for issue, state which item has what problem), list every inspected ref in inspected, and call submit_qc_report; tell the customer the QC conclusion verbatim. When agent_qc.required=true, without a submitted report this review_token cannot proceed to the next step (400 agent_qc_required); issues in the report block by default just like platform errors. ★[Customer acceptance items] review_frames / review_videos append the customer's registered specific requirements (add_acceptance_criteria) to the checklist (ids start with req:) together with acceptance_criteria: the platform's generic review does not know these requirements, so passing it does not mean they are met. Check each one with your own eyes, answer it in submit_qc_report, tell the customer the result, and let the customer confirm via confirm_acceptance_criterion.",
+  "review_storyboards": "【Hard gate ② · free】Review storyboards: forbidden terms (will be rejected by vendors and waste credits), shot duration distribution, repeated adjacent composition, character blocking drift in the same scene, emotional curve peaks/valleys, key-shot markers. ★Must run this tool before generate_frames—storyboard issues become a whole episode of wasted images once generated and cannot be fixed shot by shot. Fix per findings.action using update_shot/split_shot, then review again. ★A finding with code=forbidden_term lists every match at once (no longer only the first few) and includes `hits`: [{shot number, field, matched term}], where field is one of description / image_prompt / video_prompt / sound_effect; read that field's current value with get_shot_prompts first, then update_shot to remove the term, no need to guess word by word. ★To express lighting (backlight / side light / light and shadow), do not write it into the visual description or prompts; write it into update_shot's lighting field—it is outside the forbidden-term scan, and both video and image generation read it. ★Forbidden terms are judged with **negative semantics**: writing \"no background music\" or \"do not add music\" is not a violation (it is obeying the constraint), so no edit is needed. ★\"Action overload\" in warnings (action_overload, advisory only, non-blocking): one shot contains more independent actions than its duration can support; before video generation, use split_shot to split it or delete actions—slow motion/dissolves cannot rescue it after final cut. Ignore if multiple people each do one thing or it is a naturally continuous action chain.\n★【Agent hands-on QC · free】Returned agent_qc contains this layer's checklist and materials to inspect for this episode (materials, each with a ref; images/videos include links). The platform only checks \"exists / present / rejected\"; \"looks like / is correct / is good\" must be judged by you: download and open images, sample at least beginning/middle/end frames for videos, read text fully with tools. Answer checklist items one by one (ok / issue / na; for issue, state which item has what problem), list every inspected ref in inspected, and call submit_qc_report; tell the customer the QC conclusion verbatim. When agent_qc.required=true, without a submitted report this review_token cannot proceed to the next step (400 agent_qc_required); issues in the report block by default just like platform errors. ★[Customer acceptance items] review_frames / review_videos append the customer's registered specific requirements (add_acceptance_criteria) to the checklist (ids start with req:) together with acceptance_criteria: the platform's generic review does not know these requirements, so passing it does not mean they are met. Check each one with your own eyes, answer it in submit_qc_report, tell the customer the result, and let the customer confirm via confirm_acceptance_criterion.",
   "review_frames": "【Hard gate ③ · free】Review the shot image layer: character identity-anchor coverage (characters without character sheets will drift in shots), image-generation failure rate, orphan character variants, scene images polluted by people. ★Must run this tool before generate_videos—video generation is the most expensive step in the whole chain, and generating a whole episode of video from drifting first frames is the most typical wasted-film failure. Fix per findings.action (mostly generate_character_portraits / generate_shot_frame single-shot regeneration), then review again. ★Also reports **inter-shot seams** (warning, non-blocking): code=seam_unchained (marked continuous but never did first/last-frame chaining, or initiated chained regeneration never landed / first frame was regenerated again after chaining) / seam_broken (chained but not connected, or previous shot's last frame was regenerated after chaining); shots is the shot whose \"seam with previous shot\" has the issue. Cheapest to fix before video generation—videos are generated from first frames; if a first frame does not match the previous shot's last frame, the final cut has a jump cut. See action for fixes (whole episode chain_frames dry_run first; single seam chain_shot_from_prev / chain_shot_from).\n★【Agent hands-on QC · free】Returned agent_qc contains this layer's checklist and materials to inspect for this episode (materials, each with a ref; images/videos include links). The platform only checks \"exists / present / rejected\"; \"looks like / is correct / is good\" must be judged by you: download and open images, sample at least beginning/middle/end frames for videos, read text fully with tools. Answer checklist items one by one (ok / issue / na; for issue, state which item has what problem), list every inspected ref in inspected, and call submit_qc_report; tell the customer the QC conclusion verbatim. When agent_qc.required=true, without a submitted report this review_token cannot proceed to the next step (400 agent_qc_required); issues in the report block by default just like platform errors. ★[Customer acceptance items] review_frames / review_videos append the customer's registered specific requirements (add_acceptance_criteria) to the checklist (ids start with req:) together with acceptance_criteria: the platform's generic review does not know these requirements, so passing it does not mean they are met. Check each one with your own eyes, answer it in submit_qc_report, tell the customer the result, and let the customer confirm via confirm_acceptance_criterion. ★When there are adjacent shots, the response also carries seam_direction_advisory (advisory, does not affect review_token): existing results of the adjacent-shot direction check (subject side jump/reversed motion/shot-size jump); seams not yet observed need audit_seam_directions dry_run for a quote first, and run only after the client agrees. ★The response may carry an **unchained-seam hint** (does not affect review_token / quote / status codes): review_frames and generate_videos return unchained_seam_advisory; quote_videos returns an advisories entry with code=unchained_seams. seams lists adjacent seams (from_shot→to_shot) that chain_frames would chain but that are not chained yet: status=no_stamp (never chained) / pending (a chaining was dispatched but has not landed, or its audit has no verdict yet) / stale_anchor (the previous chaining anchor is no longer valid: the upstream shot's last frame was regenerated after chaining, so the seam must be re-chained) / join_failed (the last chaining regeneration was rejected / failed / superseded and never landed) / audit_failed (the chaining landed but the continuity audit judged it discontinuous — see seam_failed_dims in get_storyboards) / first_frame_regenerated (the downstream first frame was regenerated after the seam was reviewed, bypassing the chaining path, so the old chaining no longer counts) / not_linked (the planner judges the seam unchained but the cause is not evidenced); re-chain all of them per the plan; video_locked lists the same kind of seam where the downstream shot already has a video — chain_frames will not touch those, so fix them per shot with chain_shot_from_prev / chain_shot_from and then regenerate_shot_video (quote first). If seams is non-empty before generating videos, run chain_frames(dry_run:true) first, tell the customer the plan and price, execute after they agree, then re-run review_frames.",
   "review_assets": "【Asset-layer hard gate · free】Review character appearance images, scene images, and prop images used **on screen** in this episode—coverage and quality: on-screen character missing character sheet / missing reference image / still generating / generation failed / reference image rejected by consistency gate (hint includes rejection reason); scene missing empty-scene base plate; prop referenced by storyboard missing prop image / rejected; prop image conflicts with this shot's era/form; missing color script (warning). ★Call once after producing character sheets, reference images, scene images, and prop images; must review before generate_frames / generate_shot_frame—pass the returned review_token as **asset_review_token**, together with the storyboard-layer review_token (both are required). Only checks the assets themselves: generating shot images later will not invalidate it; changing character sheets / scene images / prop images will invalidate it, just review again (free, seconds). Explain findings to the customer one by one (code=issue type, names=which character/scene/prop, shots=affected shot numbers, action=which tool to call to fix).\n★【Agent hands-on QC · free】Returned agent_qc contains this layer's checklist and materials to inspect for this episode (materials, each with a ref; images/videos include links). The platform only checks \"exists / present / rejected\"; \"looks like / is correct / is good\" must be judged by you: download and open images, sample at least beginning/middle/end frames for videos, read text fully with tools. Answer checklist items one by one (ok / issue / na; for issue, state which item has what problem), list every inspected ref in inspected, and call submit_qc_report; tell the customer the QC conclusion verbatim. When agent_qc.required=true, without a submitted report this review_token cannot proceed to the next step (400 agent_qc_required); issues in the report block by default just like platform errors. ★[Customer acceptance items] review_frames / review_videos append the customer's registered specific requirements (add_acceptance_criteria) to the checklist (ids start with req:) together with acceptance_criteria: the platform's generic review does not know these requirements, so passing it does not mean they are met. Check each one with your own eyes, answer it in submit_qc_report, tell the customer the result, and let the customer confirm via confirm_acceptance_criterion.",
   "review_videos": "【Video and dubbing-layer hard gate · free】Review this episode's videos and dubbing—coverage and quality: missing video / still generating / generation failed / vendor rejected; first frame changed but video was generated from an old frame (video_stale); suspected hard cut at clip end, unexpected human voice in no-dialogue shots (warning); dubbed drama: dialogue shot missing dubbing, speaking character has no voice, missing narrator role; original-audio drama: shots not attached to current voice line at export time, characters without locked voice lines; missing music (warning). ★Call once after video, dubbing (generate_tts), and music are produced; must review before compose_episode, passing the returned review_token. Finer checks (requiring decoding picture/audio tracks) are available via scan_intra_shot_cuts / scan_dialogue_coverage / audit_silent_shot_audio; after assembling the final cut, run review_final_cut.\n★★【Pre-compose checks · after review_videos passes, before compose_episode】Don't hunt for problems by repeatedly exporting the whole cut and trying transitions by hand — per-shot and seam problems have dedicated tools; run them in order before stitching: ① portraits / turnarounds / scene images / shot frames were changed in this episode → first get_stale_portrait_frames(episode_id, free) to see which shots' in-use frames are still based on old assets; for shots where it is unclear which version to use, get_asset_versions(storyboard_id, free) shows each shot's versions and dependencies (unknown means insufficient evidence, not stale; regenerating is paid, so give the list to the customer and let them decide); ② scan_intra_shot_cuts(episode_id, free) finds shots where the vendor hard-cut inside a single shot; ③ per shot, recommend_trim_window(storyboard_id, free) shows a frozen tail and the effective action window (shots inside a scene group return applicable=false — skip them); apply a recommendation with trim_shot (free); ④ audit_seams(episode_id) checks editing continuity at seams between adjacent shots: by default dry_run only lists what would be audited (free); actually running it is billed by real text-token usage — tell the customer, and only after they explicitly accept pass dry_run:false + confirm_usage_billing:true; read the result from seam_review in get_storyboards; ⑤ then compose_episode. If the final cut shows jump cuts / dragging tails / seams that don't match, go back to these tools to locate the shots and fix only those — don't keep re-exporting the whole cut by trial and error. These checks are advisory, not gates; trimming and seam audits don't invalidate the review_videos token. The response's pre_compose_checks carries the same list in structured form (tool/scope/when/cost per item).\n★【Agent hands-on QC · free】Returned agent_qc contains this layer's checklist and materials to inspect for this episode (materials, each with a ref; images/videos include links). The platform only checks \"exists / present / rejected\"; \"looks like / is correct / is good\" must be judged by you: download and open images, sample at least beginning/middle/end frames for videos, read text fully with tools. Answer checklist items one by one (ok / issue / na; for issue, state which item has what problem), list every inspected ref in inspected, and call submit_qc_report; tell the customer the QC conclusion verbatim. When agent_qc.required=true, without a submitted report this review_token cannot proceed to the next step (400 agent_qc_required); issues in the report block by default just like platform errors. ★If the response includes post_production_advisory = narrative-required effects that the script marks as post-production are still unhandled (the platform will not composite them): tell the client verbatim and use get_post_production_checklist to see the list. ★[Customer acceptance items] review_frames / review_videos append the customer's registered specific requirements (add_acceptance_criteria) to the checklist (ids start with req:) together with acceptance_criteria: the platform's generic review does not know these requirements, so passing it does not mean they are met. Check each one with your own eyes, answer it in submit_qc_report, tell the customer the result, and let the customer confirm via confirm_acceptance_criterion.",
   "submit_qc_report": "Submit the QC report made by you (agent) after personally inspecting materials. Free. First call the corresponding review_* layer (script / storyboards / assets / frames / videos), open each returned agent_qc material in materials, answer every checklist item, then call this tool. ★The platform only verifies the report is **complete**, not whether your conclusions are correct: every checklist id must be answered (ok / issue / na; issue must include note explaining which item has what problem); every ref in materials must be listed in inspected (omit one and it is rejected with the missing refs named); if there are issues, set verdict=issues and list them in issues; reporting pass when issues exist will be rejected. ★Do not go through the motions: listing something in inspected without opening it means you signed off for the customer—wasted output will still be charged. ★When verdict=issues, this token cannot proceed to the next step by default: explain the problems to the customer, fix, then review + QC again; only if the customer is clearly informed and insists on proceeding as-is should the next step include acknowledge_review:true. ★suggestions (change suggestions): required at script layer when agent_qc.suggestions_required=true and issues were reported; optional for other layers. Each item includes problem + proposal; if text can be directly changed, also provide before/after; the response assigns each suggestion an id (s1…). Explain suggestions to the customer; the customer chooses accept some / execute all directly / proceed without changes (which route; autopilot = only execute priority=must suggestions that do not change the customer's original lines per execution mode: professional mode / unconfirmed plan / this step customer-controlled → explain to customer and only proceed after approval; do not decide for the customer. This step autopilot → do not ask customer; follow get_production_plan autopilot.default_decisions, recorded by the platform in get_autopilot_log). If content changes after the review, the report will be rejected; review again and QC against the new agent_qc.",
   "review_all": "(Recommended · free) Overall health check running three layers at once: script + storyboards + shot images. Use for initial assessment before work starts and recheck before delivery. ★It **does not issue review_token**—before paid steps, still run the corresponding review_script / review_storyboards / review_frames once each (the gate recognizes \"just reviewed against the current output\", not \"reviewed once sometime before\").",
-  "autofill_storyboards": "(Recommended) AI one-click fills missing fields across all episode storyboards; by default it **only fills blanks and does not overwrite existing values** (only overwrites when overwrite=true). Improves storyboard completeness; do before image generation. Async in background, postpaid text step. ★When platform contract auto-fill is enabled, it also tries to add contracts to shots **without composition contracts** (whether enabled is indicated by the note in the import receipt contract_coverage; visible headcount / subject range / shot size are judged by it for both image generation and review; with default overwrite:false, existing contracts are untouched; **overwrite:true also rewrites existing contracts**)—if auto_complete was off during import and the contract_coverage note points to this tool, call it once before image generation. If this episode already has autofill running, this run **will not queue** (prevents duplicate-click duplicate charges); newly imported shots must wait until get_autofill_status is done before calling again.",
+  "quote_autofill": "Quote: credits for autofill_storyboards. No charge, changes no data. Same basis as the completion itself: in fill-empty mode, shots with no gaps are skipped and not billed; overwrite:true bills every shot. Returns { shots_to_fill (shots that will actually call AI), shots_skipped_no_gap (shots with no gaps, skipped and not charged), estimated_points (budget upper bound; prepare balance by this), typical_points (usual cost) }—tell the customer both point figures, and after consent call autofill_storyboards with the **same** overwrite / storyboard_ids. Completion is a postpaid text step; this quote issues no quote_id and does not lock the price.",
+  "autofill_storyboards": "(Recommended) AI one-click fills missing storyboard fields across the episode (or only the shots in storyboard_ids); by default it **only fills blanks and does not overwrite existing values** (only overwrites when overwrite=true). Improves storyboard completeness; do before image generation. Async in background, postpaid text step (billed per shot that actually calls AI). ★**Quote first with quote_autofill** (same overwrite / storyboard_ids), tell the customer shots_to_fill and estimated_points / typical_points, and **run this tool only after the customer agrees**. ★When done, read the receipt from get_autofill_status (which fields were written on how many shots, how many skipped, how many still lack contracts) and tell the customer—the internal fields written here are not readable via get_storyboards, so don’t let the customer think they were charged for nothing. ★The platform already completed one round when storyboards were saved: in fill-empty mode, shots with no gaps left (all six narrative text fields present, vocal direction present on shots with dialogue, and contract present when contract auto-fill is on) are skipped with no AI call and no charge; only overwrite:true reruns the whole episode and bills every shot. ★Completion writes narrative/rhythm internal fields and composition contracts; get_storyboards / get_shot_prompts do not return these fields, so no visible change on read-back is normal—check the counts in the get_autofill_status receipt; it **does not change** image/video prompt text. ★When platform contract auto-fill is enabled, it also tries to add contracts to shots **without composition contracts** (whether enabled is indicated by the note in the import receipt contract_coverage; visible headcount / subject range / shot size are judged by it for both image generation and review; with default overwrite:false, existing contracts are untouched; **overwrite:true also rewrites existing contracts**)—if auto_complete was off during import and the contract_coverage note points to this tool, call it once before image generation. If this episode already has autofill running, this run **will not queue** (prevents duplicate-click duplicate charges); newly imported shots must wait until get_autofill_status is done before calling again. ★To trial a small range first: pass storyboard_ids (1–50 shot ids of this episode, from get_storyboards); only those shots are filled, others make no AI call and are not charged; omit = whole episode. If a completion is already running in this episode, the subset does not start: the receipt returns started=false, reason=fill batch already running; call again after get_autofill_status reaches done.",
   "enhance_shot_prompts": "(Optional enhancement · ★not required · has side effects) AI batch-enhances the whole episode's image_prompt. ★It **rewrites existing prompts** (including manually edited ones); and **must be done before generate_frames**—changing prompts after image generation makes images stale and requires paid regeneration. Not required to get a film through the pipeline; do only if the customer explicitly requests it. Async in background, postpaid text step.",
   "complete_ending_motifs": "(Optional enhancement·★not required·conditional) Complete missing visual motifs in the ending segment (graft visual echoes onto ending shots). ★Prerequisite: this drama already has motifs (none = no-op); this will **modify ending shot action/descriptions**, so do it before generating images (otherwise already-generated ending shots become stale). Backend async, billed after the text step.",
   "tail_frame_plan": "★Free read-only: which shots in this episode **need independent last frames** (shots whose terminal state ≠ initial state—large camera move/object released/state change). It uses the same criteria the platform uses for video generation; you cannot guess this yourself, so call it once after first frames are done. Returns shots_needing_last_frame(total), ready(shots with first frames already available and can get last frames now), pending_first_frame(shots that need last frames but whose first frames are not ready yet—call this tool again after first frames finish). After getting the ready list: use quote_frames + generate_frames with frame_type=last_frame to batch-fill. ★After last frames are filled (or this tool reports 0), the next step is seams: chain_frames(dry_run:true) to view plan and estimate, execute after customer confirmation, then review_frames. ★Skipping last frames will not error, but those shots have only one first-frame anchor during video generation, so the model improvises the terminal state—bad action completion, large camera moves ending back at the starting composition, etc. usually come from this. Zero cost, callable anytime.",
@@ -135,7 +136,7 @@
   "scan_dialogue_coverage": "★When the customer reports \"cut before the line finished / only half the dialogue was spoken\", run this first. Default **free, zero charge** (read-only access to persisted transcription audit results). Compares each original-voice shot's video audio track against the dialogue at word level, and reports conclusions by family: truncated (stops halfway through the line) / off_script (most spoken content is not the dialogue: improvising, repeating previous shot, only ambience) / extra_speech (dialogue is fully spoken but extra speech is added) / minor_gap (sporadic missing words, transcription error, ignore) / ok. ★Interpretation: the three reportable families have **different remedies**; do not simply lengthen every shot—measurements show 21% of shots with sufficient length still fail to finish the line, so \"too short\" is rarely the real cause (the platform already auto-raises duration by dialogue when saving storyboards). The correct fix for truncated/off_script is regenerate_shot_video (how much the vendor says is random each time; rerolling often gets the full line); if the same drama repeatedly has off_script, switch engine. ★never_audited is the most important section of this report: those shots have never been checked; \"not checked\" does not mean \"no issue\". rescan=true will run ASR comparison for them (compute-intensive, not instant; limit controls per-run cap), default does not run. TTS-dubbed shots are out of scope (TTS necessarily reads the full line; low coverage can only be transcription error). If the same episode also has \"cuts too fast\", that is a separate issue; run scan_intra_shot_cuts.",
   "update_shot": "Shot-by-shot edit: change a single storyboard shot's text content (shot size/action/dialogue/visual description/camera movement, etc.), facial expression (expression_hint), duration (duration), final cut speed (speed_factor), character binding (character_ids), voice-over marker (character_presence), and wardrobe binding (active_wardrobe_id / appearance_wardrobes). Send only fields to change; all others stay unchanged. ★Any shot edit (accepted or rejected) invalidates unconsumed quotes covering this shot; generation then returns 409 QUOTE_INVALIDATED—edit first, read back to verify, then quote. **Free** (pure text DB write). ★Changing dialogue automatically invalidates generated TTS voice-over and subtitles for this shot (tts must be regenerated); text edits do not auto-regenerate images/video. To make visuals follow text changes, edit then regen the corresponding shot. Use get_storyboards to inspect results (each shot includes character_ids / character_bindings / scene_id / prop_ids / active_wardrobe_id / appearance_wardrobes). ★★【Change clothes here, not in text】What appears in image generation is determined by **wardrobe binding**: single-character shots use active_wardrobe_id (one outfit for the whole shot); multi-character shots use appearance_wardrobes (per character). Writing \"wearing X clothes\" in description/action will not change clothes. Use list_wardrobe for wardrobe id; edits do not auto-regenerate images. To update visuals, regenerate this shot's first frame then video. ★★For native-audio shots (vendor-native audio), after changing dialogue the shot video is marked \"needs regeneration\"—because spoken lines are **baked into the video audio**. If you final-compose without regeneration, the final cut still speaks the old line (typical symptom: dialogue seems to have moved to another shot). compose_episode lists these shots under advisory `stale_video_after_edit`; correct handling is regenerate_shot_video before final compose. ★You may also directly edit this shot's four prompt bodies (image_prompt/video_prompt/first_frame_prompt/last_frame_prompt): read current values with get_shot_prompts first; do not overwrite from scratch. @char:N / @scene:M in the body are supplemental reference markers for image generation; keep them unchanged. ★Which character sheet is injected is determined by character_ids, and background by scene_id: deleting markers will not remove already-bound people from the image. To make someone not appear, edit character_ids; if they are voice-only, set character_presence to voice_only. Conversely, do not write @char markers for unbound characters—this will inject their image. ★★【If last-frame generation is blocked, edit here】If last-frame generation reports \"this shot is marked as a state change, but nowhere describes what it looks like at the end\" (TERMINAL_DESC_GATE), fix by writing **what the final image should look like** into this shot's last_frame_prompt, then regenerate the last frame. Do not retry image generation—shots of this type (action_motion_class=state_change without terminal state) historically succeed at 9.4%, versus 40.2% with terminal-state description; without adding the description, the vendor prompt has no terminal-state section, so it still draws the same as the first frame, is rejected again, and each regeneration still charges normally. terminal_desc_missing=true from get_shot_prompts is this signal.",
   "rederive_shot_contract": "Using this shot's **current** prompt body and character binding, let the platform re-derive the composition contract (how many visible people, whether the subject is a person or a partial/object, shot size). The contract itself is internal: cannot be read or edited directly. This tool only returns outcome: updated (changed and effective) / unchanged (derived result same as current) / rejected (derived result self-contradictory, not adopted, original contract kept) / no_candidate (no usable result, original contract kept) / stale (shot changed during derivation, result not written; call again after edits), plus note for next step. Insufficient credits returns 402. ★When to use: image generation/precheck/review repeatedly says the number of people is wrong, or a partial shot (hand/prop close-up) is treated as a full-person shot in generation or review, and you have already edited text or bindings—the autofiller only fills blanks and does not overwrite existing contracts; without this tool, edits will not enter the contract. ★Correct order: first update_shot to make the image text correct (who is visible, who is voice-only via character_presence voice_only, unbind anyone who should not appear from character_ids), then call this tool; calling without text edits usually returns unchanged. ★One text AI call, postpaid by token (same chain as autofill_storyboards); inform the customer before calling. ★Only changes this one contract field; does not touch dialogue/prompts/other specialized fields. Generated first/last frames are not auto-regenerated; to update visuals, regenerate frames for this shot (quote first). ★updated invalidates unconsumed frame/video quotes for this shot and the same episode; quote again afterward.",
-  "get_shot_prompts": "Read the four prompt **bodies** for one shot (image_prompt=first-frame image / video_prompt=dynamic performance / first_frame_prompt=what it looks like at start / last_frame_prompt=what it looks like at end), for direct fine-tuning then writing back with update_shot. Free. ★terminal_desc_missing=true in the response means this shot is marked as a \"state change\" but none of the three places describes the terminal state—last-frame generation will be blocked by the pre-gate. Fix by writing the end state into last_frame_prompt (do not retry image generation; retry will inevitably be rejected again and still charged). ★Also returns expression_hint (author-specified facial expression for this shot; null=unspecified, platform infers from dialogue). Edit it with update_shot. expression_inferred is the storyboard AI's inferred expression when the script did not specify expression (read-only; null=no inference, or this shot's dialogue/action/mood/description/intensity changed and the inference is expired and no longer used); it has lower priority than expression_hint—if inference is unsatisfactory, write expression_hint with update_shot to override it. ★Per-shot on demand: read only the shot you edit (episode list get_storyboards does not include prompts). ★Returned asset_tokens are supplemental reference markers in the body (@char:N / @scene:M); keep them unchanged when rewriting. Which character sheet is injected is determined by bindings (character_ids / scene_id in get_storyboards); deleting markers will not remove bound characters. Markers with #variant are the only source for that variant image. Writing a marker for an unbound character will inject that character's image. ★This is the body layer in the storyboard sheet; during image/video generation the platform also appends identity anchors, consistency constraints, and reference-image instructions (not here, and you do not need to write them). Changing prompts does not auto-regenerate images/video; to update visuals, regen the corresponding shot.",
+  "get_shot_prompts": "Read the four prompt **bodies** for one shot (image_prompt=first-frame image / video_prompt=dynamic performance / first_frame_prompt=what it looks like at start / last_frame_prompt=what it looks like at end), for direct fine-tuning then writing back with update_shot. Free. ★terminal_desc_missing=true in the response means this shot is marked as a \"state change\" but none of the three places describes the terminal state—last-frame generation will be blocked by the pre-gate. Fix by writing the end state into last_frame_prompt (do not retry image generation; retry will inevitably be rejected again and still charged). ★Also returns the current description (visual description) and lighting (shot lighting)—review_storyboards forbidden-term matches are often in description, so read it here before editing; write lighting into lighting. ★Also returns expression_hint (author-specified facial expression for this shot; null=unspecified, platform infers from dialogue). Edit it with update_shot. expression_inferred is the storyboard AI's inferred expression when the script did not specify expression (read-only; null=no inference, or this shot's dialogue/action/mood/description/intensity changed and the inference is expired and no longer used); it has lower priority than expression_hint—if inference is unsatisfactory, write expression_hint with update_shot to override it. ★Per-shot on demand: read only the shot you edit (episode list get_storyboards does not include prompts). ★Returned asset_tokens are supplemental reference markers in the body (@char:N / @scene:M); keep them unchanged when rewriting. Which character sheet is injected is determined by bindings (character_ids / scene_id in get_storyboards); deleting markers will not remove bound characters. Markers with #variant are the only source for that variant image. Writing a marker for an unbound character will inject that character's image. ★This is the body layer in the storyboard sheet; during image/video generation the platform also appends identity anchors, consistency constraints, and reference-image instructions (not here, and you do not need to write them). Changing prompts does not auto-regenerate images/video; to update visuals, regen the corresponding shot.",
   "rerender_episode": "Recompose the final cut from the current timeline (use after shot edits/cropping). **Free** (pure ffmpeg+COS), asynchronous in background.",
   "replace_shot_dialogue": "Replace a shot's dialogue timbre/voice (transcribe + clone + redub). **Synchronous call**, returns product in about 1 minute, postpaid by usage with no debt. Requires this shot to have native-audio video + character voice character sheet audio. ★This is also the cheapest fix for \"this shot did not finish the line / spoke something else\": redub the full line with cloned voice and mute the native-audio segment, **without regenerating video**—charged at TTS tier (per thousand characters), orders of magnitude cheaper than regenerate_shot_video (billed per second; over a hundred credits per second at 720p). Tradeoff: the image lip-sync was performed to the original audio, so after replacement it may mismatch; almost lossless when lips are unclear (back view/long shot/voice-over), use caution on extreme close-ups. ★Product is automatically written back to the shot (returns persisted:true), but written back ≠ included in final cut: rely on productEffective from get_dialogue_repair_status—if true, rerender_episode to recompose; if false, first perform pendingCompose as instructed, then recompose, otherwise the final cut still uses old audio. ★If this shot already has a lip-sync product, the final cut prioritizes the lip-sync version, so replacing track will not take effect. For episode batch use repair_episode_dialogue.\n★Before writeback, the product is verified (same criteria as scan_dialogue_coverage): still truncated, repeated/extra speech (extra_speech), or not the line (off_script) → **not written back**, this cost automatically refunded, and error returns which category—these shots cannot be fixed by voice replacement; use regenerate_shot_video. On success, verify:verified=false means it could not be judged (transcription failed/line too short); it is still written back, but unverified, so listen yourself.",
   "revoice_narration": "Replace the voice for a shot's **narration/voice-over** (or re-synthesize after editing narration text). **Does not regenerate video**: keeps visuals and on-screen dialogue, only mutes the narration segments, synthesizes them with the current narration voice, mixes back, then verifies. Synchronous call, charged at TTS tier (per thousand characters). ★Edit first, then call: to change voice, first set_character_voice for the \"narration\" character (or design_voice to create one); to change text, first update_shot the narration line in dialogue. ★Pure narration shots (only narration in this shot) work directly; **mixed shots** (on-screen dialogue plus narration) require sentence count to match the speech segments in the video. If not, it errors instead of guessing (a wrong guess could mute on-screen dialogue too)—then only regenerate_shot_video can fix it. ★If the dialogue field has no narration line, it errors directly; if the vendor randomly speaks in a no-dialogue shot, use replace_shot_dialogue mute_only. ★After writeback, the final cut uses it directly (no extra switch); rerender_episode is enough to recompose. If verification fails, it is not written back, and narration synthesis cost is not refunded.",