@starreel/mcp 0.1.67 → 0.1.68

package/SKILL.md

@@ -652,6 +652,16 @@ to close") tells the vendor to fit that entire sequence into each 3-second shot.
   each**: `apply_recovered_final`, or `apply_recovered_candidate` (which
   requires an explicit `target_storyboard_id` — the platform will not guess
   which shot a recovered frame belongs to). `rollback_asset_recovery` undoes one.
+- **Final-cut review (advisory, not a gate)**: `review_final_cut` — unrendered
+  shots, subtitles that won't fit on screen, characters without an identity
+  anchor, axis jumps, plus one LLM proofreading pass (that pass **costs money**;
+  send `proofread: false` to skip it). `can_export: false` only means something
+  is worth a look — the platform will **not** block `compose_episode` or
+  `get_export`. Walk the findings past the user and let them decide.
+  `audit_silent_shot_audio` catches reaction shots that should be silent but are
+  louder than the speaking shots around them (only relevant when the drama uses
+  clip audio); it is **slow** — 6–10s per shot — so run it once before the final
+  cut, not repeatedly mid-pipeline.
 - **Marketing export**: `export_sheet_compare` / `get_sheet_compare` — the
   character-sheet-vs-frame comparison sheet. **Off by default per drama**; a 403
   or `enabled: false` means the user has to switch it on in the drama settings.

package/dist/tools/produce.js

@@ -629,6 +629,25 @@ export function registerProduceTools(server, client) {
     server.tool('get_final_cut', '查某一集成片状态与下载链接。status=completed 时返回 download_url(我方 COS 直链,可直接下载)。免费。' +
         '★bgm_stale=true 表示配乐在成片之后生成/改动、尚未进成片:重新 compose_episode(免费)即可,别用 re-render。', { episode_id: z.number().int().positive() }, async ({ episode_id }) => jsonResult(await client.produceGet(`/episodes/${episode_id}/final-cut`)));
     server.tool('get_export', '查某一集导出/母版状态(成片终拼后的可下载母版)。免费。', { episode_id: z.number().int().positive() }, async ({ episode_id }) => jsonResult(await client.produceGet(`/episodes/${episode_id}/export`)));
+    server.tool('review_final_cut', '**成片体检**:导出前把整集过一遍——没出图的镜、字幕一屏放不下、人物没有身份锚、'
+        + '镜头连续性(跳轴)、以及一次错别字通读。'
+        + '\n★**建议式,不阻断**:can_export=false 只代表有 block 级问题值得先看,'
+        + '平台不会拦着你 compose_episode / get_export。把 findings 逐条讲给客户'
+        + '(code=问题类型、shots=命中的镜号、action=该调哪个工具去修),**由客户决定修还是照发**。'
+        + '\n★费用:本身免费,但默认会跑一次 LLM 错别字通读——那一步**要花钱**。'
+        + '不需要通读就传 proofread=false(更快也不扣费)。'
+        + '\n★字幕与错别字那几项的**原文**在官网体检面板里看,这里只给结论与镜号。', {
+        episode_id: z.number().int().positive(),
+        proofread: z.boolean().optional()
+            .describe('是否跑 LLM 错别字通读(默认 true,**要花钱**);传 false 跳过这一步'),
+    }, async ({ episode_id, proofread }) => jsonResult(await client.produceGet(`/episodes/${episode_id}/qc${proofread === false ? '?proofread=0' : ''}`)));
+    server.tool('audit_silent_shot_audio', '查**无台词镜里的异常发声**:本该沉默的反应镜,音轨却比同场说话镜还响——'
+        + '成片里的表现是「有人在画外乱说话」。免费。'
+        + '\n★用在**原声剧**上(use_clip_audio=true,即直接用 AI 视频自带声音那种)。'
+        + '走 TTS 配音的片子不会有这个问题。'
+        + '\n★**很慢**:每镜要完整解码音轨,实测 6~10 秒/镜(后端限并发 4),整集几十镜要等几分钟。'
+        + '别在产线中途反复调,放在成片前查一次。'
+        + '\n★shots_probe_failed 是**没测成**的镜数,不代表那些镜有问题。', { episode_id: z.number().int().positive() }, async ({ episode_id }) => jsonResult(await client.produceGet(`/episodes/${episode_id}/silent-shot-audio-audit`)));
     server.tool('get_asset_recovery', '列本集可恢复的**历史产物**(免费·只读):以前生成过、后来被覆盖掉的成片与候选帧。'
         + '\n用途:客户说「上一版那个好」「前天那张图比现在这张好」时,从这里找回来。'
         + '\n★这一步只看不改。把候选逐条讲给客户,**客户逐条点头**之后再 apply。', { episode_id: z.number().int().positive() }, async ({ episode_id }) => jsonResult(await client.produceGet(`/episodes/${episode_id}/asset-recovery`)));

package/openapi.json

@@ -2,8 +2,8 @@
   "openapi": "3.1.0",
   "info": {
     "title": "StarReel Production API",
-    "version": "0.1.67",
-    "description": "Turn a script into a finished, downloadable short-drama episode over REST.\n\nPipeline: script → AI rewrite → cast/scenes/props extraction → portraits & sheets → storyboards → keyframes → video shots → TTS → final cut (.mp4).\n\n**Billing is prepaid and agent-safe**: big-ticket stages are quote-then-generate (`quote_*` returns a `quote_id`; for video, quote == actual charge). Insufficient balance returns 402 — nothing half-runs and the account never goes negative.\n\nAuth: exchange your API key at `POST /v1/agent/token` for a 15-minute bearer token.\n\nGenerated from the @starreel/mcp v0.1.67 tool surface (operationIds match MCP tool names 1:1)."
+    "version": "0.1.68",
+    "description": "Turn a script into a finished, downloadable short-drama episode over REST.\n\nPipeline: script → AI rewrite → cast/scenes/props extraction → portraits & sheets → storyboards → keyframes → video shots → TTS → final cut (.mp4).\n\n**Billing is prepaid and agent-safe**: big-ticket stages are quote-then-generate (`quote_*` returns a `quote_id`; for video, quote == actual charge). Insufficient balance returns 402 — nothing half-runs and the account never goes negative.\n\nAuth: exchange your API key at `POST /v1/agent/token` for a 15-minute bearer token.\n\nGenerated from the @starreel/mcp v0.1.68 tool surface (operationIds match MCP tool names 1:1)."
   },
   "servers": [
     {
@@ -4925,6 +4925,53 @@
         }
       }
     },
+    "/episodes/{episode_id}/qc": {
+      "get": {
+        "operationId": "review_final_cut",
+        "summary": "**成片体检**:导出前把整集过一遍——没出图的镜、字幕一屏放不下、人物没有身份锚、镜头连续性(跳轴)、以及一次错别字通读",
+        "description": "**成片体检**:导出前把整集过一遍——没出图的镜、字幕一屏放不下、人物没有身份锚、镜头连续性(跳轴)、以及一次错别字通读。\n★**建议式,不阻断**:can_export=false 只代表有 block 级问题值得先看,平台不会拦着你 compose_episode / get_export。把 findings 逐条讲给客户(code=问题类型、shots=命中的镜号、action=该调哪个工具去修),**由客户决定修还是照发**。\n★费用:本身免费,但默认会跑一次 LLM 错别字通读——那一步**要花钱**。不需要通读就传 proofread=false(更快也不扣费)。\n★字幕与错别字那几项的**原文**在官网体检面板里看,这里只给结论与镜号。",
+        "tags": [
+          "episodes"
+        ],
+        "parameters": [
+          {
+            "name": "episode_id",
+            "in": "path",
+            "required": true,
+            "schema": {
+              "type": "integer",
+              "exclusiveMinimum": 0
+            }
+          }
+        ],
+        "responses": {
+          "200": {
+            "description": "StarReel envelope",
+            "content": {
+              "application/json": {
+                "schema": {
+                  "type": "object",
+                  "properties": {
+                    "code": {
+                      "type": "integer"
+                    },
+                    "message": {
+                      "type": "string"
+                    },
+                    "data": {
+                      "description": "Operation result payload"
+                    }
+                  }
+                }
+              }
+            }
+          },
+          "402": {
+            "description": "Insufficient prepaid balance (never overdrafts); response carries `needed` points"
+          }
+        }
+      }
+    },
     "/episodes/{episode_id}/re-render": {
       "post": {
         "operationId": "rerender_episode",
@@ -5759,6 +5806,53 @@
         }
       }
     },
+    "/episodes/{episode_id}/silent-shot-audio-audit": {
+      "get": {
+        "operationId": "audit_silent_shot_audio",
+        "summary": "查**无台词镜里的异常发声**:本该沉默的反应镜,音轨却比同场说话镜还响——成片里的表现是「有人在画外乱说话」",
+        "description": "查**无台词镜里的异常发声**:本该沉默的反应镜,音轨却比同场说话镜还响——成片里的表现是「有人在画外乱说话」。免费。\n★用在**原声剧**上(use_clip_audio=true,即直接用 AI 视频自带声音那种)。走 TTS 配音的片子不会有这个问题。\n★**很慢**:每镜要完整解码音轨,实测 6~10 秒/镜(后端限并发 4),整集几十镜要等几分钟。别在产线中途反复调,放在成片前查一次。\n★shots_probe_failed 是**没测成**的镜数,不代表那些镜有问题。",
+        "tags": [
+          "episodes"
+        ],
+        "parameters": [
+          {
+            "name": "episode_id",
+            "in": "path",
+            "required": true,
+            "schema": {
+              "type": "integer",
+              "exclusiveMinimum": 0
+            }
+          }
+        ],
+        "responses": {
+          "200": {
+            "description": "StarReel envelope",
+            "content": {
+              "application/json": {
+                "schema": {
+                  "type": "object",
+                  "properties": {
+                    "code": {
+                      "type": "integer"
+                    },
+                    "message": {
+                      "type": "string"
+                    },
+                    "data": {
+                      "description": "Operation result payload"
+                    }
+                  }
+                }
+              }
+            }
+          },
+          "402": {
+            "description": "Insufficient prepaid balance (never overdrafts); response carries `needed` points"
+          }
+        }
+      }
+    },
     "/episodes/{episode_id}/storyboards": {
       "get": {
         "operationId": "get_storyboards",

package/package.json

@@ -1,6 +1,6 @@
 {
   "name": "@starreel/mcp",
-  "version": "0.1.67",
+  "version": "0.1.68",
   "mcpName": "ai.starreel/starreel",
   "description": "StarReel MCP server — drive the AI short-drama production pipeline (script → storyboards → frames → video → final cut) from Claude, Cursor, or any MCP client",
   "license": "MIT",

package/server.json

@@ -2,13 +2,13 @@
   "$schema": "https://static.modelcontextprotocol.io/schemas/2025-09-29/server.schema.json",
   "name": "ai.starreel/starreel",
   "description": "Turn a script into a finished, downloadable short-drama episode — via MCP or REST.",
-  "version": "0.1.67",
+  "version": "0.1.68",
   "websiteUrl": "https://starreel.ai",
   "packages": [
     {
       "registryType": "npm",
       "identifier": "@starreel/mcp",
-      "version": "0.1.67",
+      "version": "0.1.68",
       "transport": {
         "type": "stdio"
       },