openclaw-voc-skill/video-creation/transcript-to-video/api-config.json

{
  "name": "transcript-to-video",
  "displayName": "閫愬瓧绋緼I瑙嗛鐢熸垚宸ヤ綔娴?,
  "description": "杈撳叆閫愬瓧绋挎枃鏈紝AI鑷姩鎷嗗垎闀滆剼鏈紝鎵归噺璋冪敤鍗虫ⅵ瑙嗛鐢熸垚锛岃緭鍑鸿棰戠墖娈礥RL鍒楄〃+SRT瀛楀箷鏂囦欢銆傛敮鎸佹枃鐢熻棰戝拰鍏堝浘鍚庤棰戜袱绉嶆ā寮忋€?,
  "category": "video-creation",
  "version": "1.0.0",
  "type": "orchestration",

  "parameters": {
    "type": "object",
    "required": ["transcript"],
    "properties": {
      "transcript": {
        "type": "string",
        "description": "閫愬瓧绋垮叏鏂囷紙涓枃锛夛紝Agent灏嗚嚜鍔ㄦ媶鍒嗕负鍒嗛暅娈?
      },
      "segmentDuration": {
        "type": "number",
        "description": "姣忔鐩爣鏃堕暱锛堢锛夛紝褰卞搷鍒嗛暅鎷嗗垎绮掑害",
        "default": 15
      },
      "videoModel": {
        "type": "string",
        "enum": ["720p", "1080p", "pro"],
        "description": "瑙嗛鐢熸垚妯″瀷璐ㄩ噺绛夌骇",
        "default": "720p"
      },
      "videoMethod": {
        "type": "string",
        "enum": ["text2video", "img2video"],
        "description": "text2video=绾枃鐢熻棰戯紙蹇€燂級锛宨mg2video=鍏堢敓鍥惧啀鍥剧敓瑙嗛锛堟晥鏋滄洿鍙帶锛?,
        "default": "text2video"
      },
      "aspectRatio": {
        "type": "string",
        "enum": ["16:9", "9:16", "1:1", "4:3", "3:4", "21:9"],
        "description": "瑙嗛鐢婚潰姣斾緥",
        "default": "16:9"
      },
      "frames": {
        "type": "integer",
        "enum": [121, 241],
        "description": "瑙嗛甯ф暟锛?21=5绉掞紝241=10绉?,
        "default": 121
      }
    }
  },

  "pipeline": [
    {
      "step": 1,
      "name": "AI鍒嗛暅鑴氭湰鐢熸垚",
      "type": "ai",
      "description": "Agent灏嗛€愬瓧绋挎媶鍒嗕负鍒嗛暅JSON鍜孲RT瀛楀箷",
      "aiConfig": {
        "systemPrompt": "浣犳槸涓撲笟鐨勮棰戝垎闀滆剼鏈紪鍐欎笓瀹躲€傚皢閫愬瓧绋挎媶鍒嗕负鍒嗛暅娈碉紝姣忔绾?0-45涓腑鏂囧瓧锛堢害15绉掕閫燂級銆備负姣忔鐢熸垚鑻辨枃瑙嗛鐢熸垚鎻愮ず璇嶏紙prompt锛夛紝鎻忚堪鐢婚潰鍐呭銆侀暅澶寸被鍨嬨€佹皼鍥村拰鐢昏川銆?,
        "userPromptTemplate": "璇峰皢浠ヤ笅閫愬瓧绋挎媶鍒嗕负鍒嗛暅鑴氭湰锛屾瘡娈电害${segmentDuration}绉掋€俓n\n閫愬瓧绋匡細\n${transcript}\n\n杈撳嚭JSON鏁扮粍鏍煎紡锛歕n[{\"id\":\"S-01\",\"act\":\"娈佃惤鍚峔",\"narration\":\"涓枃鏃佺櫧鍘熸枃\",\"type\":\"ai-gen\",\"prompt\":\"English video generation prompt with scene, camera, mood, quality keywords\",\"duration\":\"15s\"}]\n\n鍚屾椂杈撳嚭SRT瀛楀箷鏂囦欢鍐呭銆?,
        "outputFormat": {
          "storyboard": "Array<{id, act, narration, type, prompt, duration}>",
          "srt": "String (SRT format subtitle content)"
        },
        "temperature": 0.3
      },
      "output": {
        "storyboard": "鍒嗛暅JSON鏁扮粍",
        "srt": "SRT瀛楀箷鏂囦欢鍐呭"
      }
    },
    {
      "step": 2,
      "name": "鎵归噺鐢熸垚瑙嗛绱犳潗",
      "type": "api",
      "description": "瀵规瘡涓垎闀滄骞惰璋冪敤鍗虫ⅵ瑙嗛鐢熸垚API",
      "forEach": "storyboard",
      "forEachAs": "segment",
      "parallel": {
        "maxConcurrent": 5,
        "delayMs": 1000
      },
      "branches": {
        "text2video": {
          "condition": "videoMethod === 'text2video'",
          "steps": [
            {
              "id": "2a",
              "name": "鏂囩敓瑙嗛",
              "skill": "jimeng/jimeng-video-v3-720p",
              "skillVariants": {
                "720p": "jimeng/jimeng-video-v3-720p",
                "1080p": "jimeng/jimeng-video-v3-1080p",
                "pro": "jimeng/jimeng-video-v3-pro"
              },
              "input": {
                "prompt": "${segment.prompt}",
                "method": "1",
                "frames": "${frames}",
                "config": {
                  "aspect_ratio": "${aspectRatio}"
                }
              },
              "output": "workIds[${segment.id}]"
            }
          ]
        },
        "img2video": {
          "condition": "videoMethod === 'img2video'",
          "steps": [
            {
              "id": "2b-1",
              "name": "鏂囩敓鍥撅紙棣栧抚锛?,
              "skill": "jimeng/jimeng-img-v4",
              "input": {
                "prompt": "${segment.prompt}",
                "sizeDate": {
                  "width": "${aspectRatio === '16:9' ? 2560 : aspectRatio === '9:16' ? 1440 : 2048}",
                  "height": "${aspectRatio === '16:9' ? 1440 : aspectRatio === '9:16' ? 2560 : 2048}"
                }
              },
              "output": "imageWorkIds[${segment.id}]"
            },
            {
              "id": "2b-2",
              "name": "杞鍥剧墖缁撴灉",
              "skill": "jimeng/jimeng-task-query",
              "dependsOn": ["2b-1"],
              "input": {
                "workId": "${imageWorkIds[segment.id]}",
                "routerName": "getImgV4"
              },
              "pollingConfig": {
                "intervalMs": 3000,
                "maxAttempts": 30,
                "completionCondition": "data.isFinish === true"
              }
            },
            {
              "id": "2b-3",
              "name": "鑾峰彇鍥剧墖URL",
              "skill": "jimeng/jimeng-work-result",
              "dependsOn": ["2b-2"],
              "input": {
                "workId": "${imageWorkIds[segment.id]}"
              },
              "output": "imageUrls[${segment.id}]",
              "responseExtract": "images[0]"
            },
            {
              "id": "2b-4",
              "name": "鍥剧敓瑙嗛锛堥甯э級",
              "skill": "jimeng/jimeng-video-v3-720p",
              "dependsOn": ["2b-3"],
              "input": {
                "prompt": "${segment.prompt}",
                "method": "2",
                "frames": "${frames}",
                "config": {
                  "image_urls": ["${imageUrls[segment.id]}"]
                }
              },
              "output": "workIds[${segment.id}]"
            }
          ]
        }
      }
    },
    {
      "step": 3,
      "name": "杞瑙嗛鐢熸垚缁撴灉",
      "type": "api",
      "forEach": "workIds",
      "forEachAs": "workId",
      "parallel": {
        "maxConcurrent": 10,
        "delayMs": 500
      },
      "steps": [
        {
          "id": "3a",
          "name": "杞浠诲姟鐘舵€?,
          "skill": "jimeng/jimeng-task-query",
          "input": {
            "workId": "${workId}",
            "routerName": "${videoModel === '720p' ? 'getVideoV3_720p' : videoModel === '1080p' ? 'getVideoV3_1080p' : 'getVideoV3_Pro'}"
          },
          "pollingConfig": {
            "intervalMs": 5000,
            "maxAttempts": 60,
            "completionCondition": "data.isFinish === true",
            "timeoutMs": 300000
          }
        },
        {
          "id": "3b",
          "name": "鑾峰彇瑙嗛URL",
          "skill": "jimeng/jimeng-work-result",
          "dependsOn": ["3a"],
          "input": {
            "workId": "${workId}"
          },
          "output": "videoResults[${segmentId}]",
          "responseExtract": {
            "videoUrl": "videos[0]",
            "images": "images"
          }
        }
      ]
    },
    {
      "step": 4,
      "name": "缁撴灉姹囨€?,
      "type": "compute",
      "logic": "aggregateResults(storyboard, srt, videoResults)",
      "algorithm": {
        "merge": "灏嗗垎闀淛SON涓庤棰慤RL鍚堝苟",
        "stats": "缁熻鎴愬姛/澶辫触鏁伴噺",
        "failedRetry": "澶辫触鐨勫垎闀滄鏍囪涓篺ailedIds锛屽彲閲嶈瘯"
      },
      "output": "finalOutput"
    },
    {
      "step": 5,
      "name": "鍙€夛細涓€閿垚鐗?,
      "type": "api",
      "optional": true,
      "condition": "user requests quick compilation && successCount <= 50 && totalDuration <= 30",
      "skill": "video-creation/quickly-video-create",
      "input": {
        "material_list": "${videoResults.map(r => ({ type: 'video', value: r.videoUrl }))}",
        "tags": "${extractKeywords(transcript)}",
        "proportion": "${aspectRatio === '9:16' ? '9:16' : aspectRatio === '1:1' ? '1:1' : '9:16'}",
        "video_duration": { "min": 15, "max": 30 },
        "pre_id": "${Date.now().toString()}",
        "ai_voice": 1,
        "ai_bgm": 1,
        "ai_subtitle": 1
      },
      "output": "quicklyTaskId",
      "downstream": "video-creation/quickly-video-query"
    }
  ],

  "response": {
    "type": "object",
    "properties": {
      "storyboard": {
        "type": "array",
        "description": "鍒嗛暅鑴氭湰JSON",
        "items": {
          "type": "object",
          "properties": {
            "id": { "type": "string" },
            "act": { "type": "string" },
            "narration": { "type": "string" },
            "prompt": { "type": "string" },
            "duration": { "type": "string" },
            "videoUrl": { "type": "string", "description": "鐢熸垚鐨勮棰慤RL锛堟案涔呮湁鏁堬級" },
            "imageUrl": { "type": "string", "description": "鐢熸垚鐨勯甯у浘URL锛坕mg2video妯″紡锛? }
          }
        }
      },
      "srt": {
        "type": "string",
        "description": "SRT瀛楀箷鏂囦欢鍐呭"
      },
      "segments": {
        "type": "array",
        "description": "瑙嗛鐗囨姹囨€?,
        "items": {
          "type": "object",
          "properties": {
            "id": { "type": "string" },
            "videoUrl": { "type": "string" },
            "duration": { "type": "number" },
            "prompt": { "type": "string" }
          }
        }
      },
      "totalSegments": { "type": "integer" },
      "successCount": { "type": "integer" },
      "failedCount": { "type": "integer" },
      "failedIds": { "type": "array", "items": { "type": "string" } }
    }
  },

  "relatedSkills": [
    "jimeng/jimeng-video-v3-720p",
    "jimeng/jimeng-video-v3-1080p",
    "jimeng/jimeng-video-v3-pro",
    "jimeng/jimeng-img-v4",
    "jimeng/jimeng-task-query",
    "jimeng/jimeng-work-result",
    "video-creation/quickly-video-create",
    "video-creation/quickly-video-query"
  ],

  "pricing": {
    "text2video_720p_5s": {
      "perSegment": 1.4,
      "unit": "CNY",
      "formula": "0.28鍏?绉?脳 5绉?
    },
    "text2video_720p_10s": {
      "perSegment": 2.8,
      "unit": "CNY",
      "formula": "0.28鍏?绉?脳 10绉?
    },
    "img2video_720p_5s": {
      "perSegment": 1.62,
      "unit": "CNY",
      "formula": "0.22鍏?鍥? + 0.28鍏?绉捗?绉?瑙嗛)"
    },
    "example_71segments_5s": {
      "text2video": "鈮?9鍏?,
      "img2video": "鈮?15鍏?
    }
  },

  "limitations": {
    "videoDuration": "浠呮敮鎸?绉?121甯?鎴?0绉?241甯?锛屾棤娉曠簿纭尮閰峉RT鏃堕棿鎴?,
    "noTTS": "鏃燭TS閰嶉煶鑳藉姏锛岄渶鐢ㄦ埛鍦ㄥ壀鏄犱腑閰嶉煶",
    "noAutoStitch": "鏃犺嚜鍔ㄦ嫾鎺ラ暱瑙嗛鑳藉姏锛宷uickly鏈€闀?0绉?,
    "parallelLimit": "鍗虫ⅵAPI寤鸿鈮?骞惰锛岄伩鍏嶉鐜囬檺鍒?
  },

  "futureExtensions": [
    {
      "name": "tts-voiceover",
      "description": "TTS閰嶉煶Skill锛氭寜閫愬瓧绋跨敓鎴愯闊抽煶棰?,
      "status": "planned"
    },
    {
      "name": "video-stitch",
      "description": "瑙嗛鎷兼帴Skill锛欶Fmpeg鏈嶅姟绔寜SRT鏃堕棿杞寸簿纭嫾鎺?,
      "status": "planned"
    },
    {
      "name": "seeddance-video",
      "description": "seeddance2.0妯″瀷鎺ュ叆锛氭洿楂樿川閲忚棰戠敓鎴?,
      "status": "planned"
    }
  ],

  "tokenConfig": {
    "type": "body",
    "tokenField": "token",
    "format": "Bearer r:{{jimengToken}}",
    "configFile": "~/.openclaw/voc-credentials.json",
    "currentToken": "Bearer r:f0333969e312a40e4703e8fe4ed1c600",
    "resolutionOrder": ["configFile", "currentToken"],
    "onMissing": {
      "action": "showPaymentQR",
      "qrCodeUrl": "https://pwa.fmode.cn/apig-pay.html?authid=T0iOotHcDX&fun_id=HOkkX72PMF",
      "paymentCheckEndpoint": "https://server.fmode.cn/api/apig/getApig",
      "pollingIntervalMs": 3000,
      "pollingTimeoutMs": 300000,
      "onPaymentSuccess": "retrySkillWithNewToken",
      "title": "鎵爜寮€閫氬嵆姊I瑙嗛鐢熸垚鏈嶅姟",
      "message": "璇锋壂鎻忎笅鏂逛簩缁寸爜瀹屾垚鏀粯锛屾敮浠樻垚鍔熷悗灏嗚嚜鍔ㄨ幏鍙?Token 骞剁户缁墽琛屻€?
    },
    "onBalanceInsufficient": {
      "action": "showPaymentQR",
      "qrCodeUrl": "https://pwa.fmode.cn/apig-pay.html?authid=T0iOotHcDX&fun_id=HOkkX72PMF",
      "paymentCheckEndpoint": "https://server.fmode.cn/api/apig/getApig",
      "pollingIntervalMs": 3000,
      "pollingTimeoutMs": 300000,
      "onPaymentSuccess": "retrySkillWithNewToken",
      "title": "鍗虫ⅵAI Token 浣欓涓嶈冻锛岃鎵爜鍏呭€?,
      "message": "褰撳墠 Token 浣欓涓嶈冻锛岃鎵弿浜岀淮鐮佸厖鍊硷紝鏀粯瀹屾垚鍚庡皢鑷姩缁х画鎵ц銆?
    }
  },

  "errorHandling": {
    "balanceInsufficient": {
      "conditions": [
        { "responseField": "code", "operator": "in", "value": [-2, -3, -10, 402, 429] },
        { "responseField": "msg", "operator": "contains", "value": ["浣欓涓嶈冻", "insufficient", "balance", "quota"] }
      ],
      "matchMode": "any",
      "trigger": "tokenConfig.onBalanceInsufficient"
    },
    "segmentFailed": {
      "action": "markFailed",
      "description": "鍗曚釜鍒嗛暅娈电敓鎴愬け璐ユ椂鏍囪涓篺ailed锛屼笉闃诲鍏朵粬娈碉紝鏈€鍚庢眹鎬诲け璐ュ垪琛ㄤ緵閲嶈瘯"
    }
  },

  "timeout": 600000,
  "retry": {
    "maxAttempts": 2,
    "delay": 3000,
    "backoffMultiplier": 2
  }
}