{
  "apiVersion": "yingce.plugin/v2",
  "id": "openai-responses",
  "name": "OpenAI Responses",
  "version": "2.0.0",
  "author": "OpenAI / 影策",
  "description": "OpenAI Responses 独立请求协议插件。",
  "documentation": "# OpenAI Responses\n\n该目录是独立官方协议插件源码。后端从生成的 `openai-responses.yingce-plugin` 包加载，不依赖系统内置 `host:` 适配器。\n\n完整接口见 [docs/interface.md](docs/interface.md)。\n\n---\n\n# OpenAI Responses 接口字段\n\n## 协议身份\n\n- 插件 ID：`openai-responses`。\n- Provider ID：`openai-response`。\n- 能力：`text`。\n- 默认 Base URL：`https://api.openai.com`。\n- 鉴权驱动：`bearer`。\n- 创建：`POST /responses`。\n- Agent：`POST /responses`。\n- 生命周期：同步响应。\n\n## 配置字段\n\n| 字段 | 类型 | 必填 | 含义 |\n| --- | --- | --- | --- |\n| `apiKey` | secret | 是 | API Key |\n\n## 统一字段映射\n\n| 统一字段 | 类型 | 必填 | 上游映射 | 说明 |\n| --- | --- | --- | --- | --- |\n| `model` | string | 是 | `model` | 上游模型 ID。 |\n| `messages` | message[] | 是 | `provider message container` | 包含历史消息和当前用户输入。 |\n| `instructions` | string | 否 | `system/instructions` | 系统指令。 |\n| `temperature` | number | 否 | `temperature` | 采样温度。 |\n| `top_p` | number | 否 | `top_p` | 核采样参数。 |\n| `max_tokens` | integer | 否 | `max_tokens/max_output_tokens` | 最大输出 token。 |\n| `tools` | array | 否 | `tools/toolConfig` | 工具定义。 |\n| `tool_choice` | object|string | 否 | `tool_choice` | 工具选择策略。 |\n| `response_format` | object | 否 | `response_format/text` | 结构化输出配置。 |\n| `stream` | boolean | 否 | `stream` | 流式开关；后台任务当前以最终响应归一。 |\n\n## 上游请求模板逐字段清单\n\n下表由插件请求模板生成，覆盖 body、query、headers 和 multipart 文件声明中的每个字段。\n\n| 上游位置 | 值或转换表达式 |\n| --- | --- |\n| `create.method` | `\"POST\"` |\n| `create.path` | `\"/responses\"` |\n| `create.contentType` | `\"application/json\"` |\n| `create.body.model` | `{\"$ref\":\"request.model\"}` |\n| `create.body.input` | `{\"$filter\":{\"from\":{\"$ref\":\"request.messages\"},\"as\":\"message\",\"where\":{\"$ne\":[{\"$ref\":\"message.role\"},\"system\"]}}}` |\n| `create.body.instructions` | `{\"$omitEmpty\":{\"$ref\":\"request.instructions\"}}` |\n| `create.body.temperature` | `{\"$omitEmpty\":{\"$ref\":\"request.providerOptions.openai-response.temperature\"}}` |\n| `create.body.top_p` | `{\"$omitEmpty\":{\"$ref\":\"request.providerOptions.openai-response.top_p\"}}` |\n| `create.body.max_output_tokens` | `{\"$omitEmpty\":{\"$coalesce\":[{\"$ref\":\"request.extra.max_output_tokens\"},{\"$ref\":\"request.providerOptions.openai-response.max_output_tokens\"}]}}` |\n| `create.body.tools` | `{\"$omitEmpty\":{\"$ref\":\"request.providerOptions.openai-response.tools\"}}` |\n| `create.body.tool_choice` | `{\"$omitEmpty\":{\"$ref\":\"request.providerOptions.openai-response.tool_choice\"}}` |\n| `create.body.text` | `{\"$omitEmpty\":{\"$ref\":\"request.providerOptions.openai-response.text\"}}` |\n| `create.body.reasoning` | `{\"$omitEmpty\":{\"$ref\":\"request.providerOptions.openai-response.reasoning\"}}` |\n| `create.body.previous_response_id` | `{\"$omitEmpty\":{\"$ref\":\"request.providerOptions.openai-response.previous_response_id\"}}` |\n| `create.body.store` | `{\"$omitEmpty\":{\"$ref\":\"request.providerOptions.openai-response.store\"}}` |\n| `create.body.metadata` | `{\"$omitEmpty\":{\"$ref\":\"request.providerOptions.openai-response.metadata\"}}` |\n| `create.body.stream` | `{\"$omitEmpty\":{\"$ref\":\"request.providerOptions.openai-response.stream\"}}` |\n| `create.body.truncation` | `{\"$omitEmpty\":{\"$ref\":\"request.providerOptions.openai-response.truncation\"}}` |\n| `create.body.user` | `{\"$omitEmpty\":{\"$ref\":\"request.providerOptions.openai-response.user\"}}` |\n| `agent.method` | `\"POST\"` |\n| `agent.path` | `\"/responses\"` |\n| `agent.contentType` | `\"application/json\"` |\n| `agent.body` | `{\"$merge\":[{\"$ref\":\"request.extra.agent.responses\"},{\"model\":{\"$ref\":\"request.model\"}}]}` |\n\n## Provider 扩展键\n\n- `providerOptions.openai-response.max_output_tokens`\n- `providerOptions.openai-response.metadata`\n- `providerOptions.openai-response.previous_response_id`\n- `providerOptions.openai-response.reasoning`\n- `providerOptions.openai-response.store`\n- `providerOptions.openai-response.stream`\n- `providerOptions.openai-response.temperature`\n- `providerOptions.openai-response.text`\n- `providerOptions.openai-response.tool_choice`\n- `providerOptions.openai-response.tools`\n- `providerOptions.openai-response.top_p`\n- `providerOptions.openai-response.truncation`\n- `providerOptions.openai-response.user`\n\n动态模型或工作流允许使用文档声明的完整 `parameters/input/extra_body` 对象；该对象是协议本身的开放 schema，不会被宿主裁剪。\n\n## 响应映射逐字段清单\n\n| 映射位置 | 上游路径或转换表达式 |\n| --- | --- |\n| `response.status` | `\"succeeded\"` |\n| `response.textPaths[0]` | `\"output_text\"` |\n| `response.reasoningPaths[0]` | `\"reasoning.summary.0.text\"` |\n| `response.usage` | `{\"$ref\":\"response.usage\"}` |\n| `response.errorPaths[0]` | `\"error.code\"` |\n| `response.messagePaths[0]` | `\"error.message\"` |\n| `agentResponse.textPaths[0]` | `\"output_text\"` |\n| `agentResponse.reasoningPaths[0]` | `\"reasoning.summary.0.text\"` |\n| `agentResponse.toolCallsPath` | `\"output\"` |\n| `agentResponse.toolCallIdPaths[0]` | `\"call_id\"` |\n| `agentResponse.toolCallIdPaths[1]` | `\"id\"` |\n| `agentResponse.toolCallNamePaths[0]` | `\"name\"` |\n| `agentResponse.toolCallArgumentsPaths[0]` | `\"arguments\"` |\n\n## 响应与错误\n\n插件把上游 task/status/text/media/usage 映射为统一结果。临时媒体 URL 标记为 ephemeral，由宿主立即下载持久化。HTTP 错误、业务 code 和 error object 保持失败语义，不包装成成功。\n\n## 兼容边界\n\n该包只代表上述线协议 profile；同一品牌的其他 endpoint、云区域或网关包装必须使用独立插件，不能根据模型名猜测。\n\n<!-- YINGCE_MANIFEST_CONTRACT_START -->\n## Manifest 完整接口定义\n\n以下 JSON 与插件包内实际 `manifest.json` 逐字段一致，覆盖插件身份、权限、配置、鉴权、参数、校验、创建、Agent、查询、取消、结果下载、响应和 Agent 响应映射。`documentation` 字段的值就是当前完整文档；为避免文档在自身内部无限递归，JSON 中仅用等义占位文本表示正文。\n\n```json\n{\n  \"apiVersion\": \"yingce.plugin/v2\",\n  \"id\": \"openai-responses\",\n  \"name\": \"OpenAI Responses\",\n  \"version\": \"2.0.0\",\n  \"author\": \"OpenAI / 影策\",\n  \"description\": \"OpenAI Responses 独立请求协议插件。\",\n  \"documentation\": \"<当前插件的完整 documentation，由 README.md 与 docs/interface.md 拼接而成；为避免 JSON 递归，此处不重复展开正文。>\",\n  \"permissions\": [\n    \"generation.run\",\n    \"media.read\"\n  ],\n  \"configuration\": {\n    \"fields\": [\n      {\n        \"name\": \"apiKey\",\n        \"type\": \"secret\",\n        \"label\": \"API Key\",\n        \"required\": true\n      }\n    ]\n  },\n  \"contributes\": {\n    \"providers\": [\n      {\n        \"id\": \"openai-response\",\n        \"label\": \"OpenAI Responses\",\n        \"capabilities\": [\n          \"text\"\n        ],\n        \"scopes\": [\n          \"admin.system-channel\",\n          \"user.custom-channel\",\n          \"canvas\",\n          \"creation\",\n          \"agent\"\n        ],\n        \"baseUrl\": \"https://api.openai.com\",\n        \"requiresPublicMediaUrls\": false,\n        \"auth\": {\n          \"type\": \"bearer\",\n          \"field\": \"apiKey\"\n        },\n        \"parameters\": [\n          {\n            \"name\": \"model\",\n            \"type\": \"string\",\n            \"required\": true,\n            \"mapping\": \"model\",\n            \"description\": \"上游模型 ID。\"\n          },\n          {\n            \"name\": \"messages\",\n            \"type\": \"message[]\",\n            \"required\": true,\n            \"mapping\": \"provider message container\",\n            \"description\": \"包含历史消息和当前用户输入。\"\n          },\n          {\n            \"name\": \"instructions\",\n            \"type\": \"string\",\n            \"required\": false,\n            \"mapping\": \"system/instructions\",\n            \"description\": \"系统指令。\"\n          },\n          {\n            \"name\": \"temperature\",\n            \"type\": \"number\",\n            \"required\": false,\n            \"mapping\": \"temperature\",\n            \"description\": \"采样温度。\"\n          },\n          {\n            \"name\": \"top_p\",\n            \"type\": \"number\",\n            \"required\": false,\n            \"mapping\": \"top_p\",\n            \"description\": \"核采样参数。\"\n          },\n          {\n            \"name\": \"max_tokens\",\n            \"type\": \"integer\",\n            \"required\": false,\n            \"mapping\": \"max_tokens/max_output_tokens\",\n            \"description\": \"最大输出 token。\"\n          },\n          {\n            \"name\": \"tools\",\n            \"type\": \"array\",\n            \"required\": false,\n            \"mapping\": \"tools/toolConfig\",\n            \"description\": \"工具定义。\"\n          },\n          {\n            \"name\": \"tool_choice\",\n            \"type\": \"object|string\",\n            \"required\": false,\n            \"mapping\": \"tool_choice\",\n            \"description\": \"工具选择策略。\"\n          },\n          {\n            \"name\": \"response_format\",\n            \"type\": \"object\",\n            \"required\": false,\n            \"mapping\": \"response_format/text\",\n            \"description\": \"结构化输出配置。\"\n          },\n          {\n            \"name\": \"stream\",\n            \"type\": \"boolean\",\n            \"required\": false,\n            \"mapping\": \"stream\",\n            \"description\": \"流式开关；后台任务当前以最终响应归一。\"\n          }\n        ],\n        \"create\": {\n          \"method\": \"POST\",\n          \"path\": \"/responses\",\n          \"contentType\": \"application/json\",\n          \"body\": {\n            \"model\": {\n              \"$ref\": \"request.model\"\n            },\n            \"input\": {\n              \"$filter\": {\n                \"from\": {\n                  \"$ref\": \"request.messages\"\n                },\n                \"as\": \"message\",\n                \"where\": {\n                  \"$ne\": [\n                    {\n                      \"$ref\": \"message.role\"\n                    },\n                    \"system\"\n                  ]\n                }\n              }\n            },\n            \"instructions\": {\n              \"$omitEmpty\": {\n                \"$ref\": \"request.instructions\"\n              }\n            },\n            \"temperature\": {\n              \"$omitEmpty\": {\n                \"$ref\": \"request.providerOptions.openai-response.temperature\"\n              }\n            },\n            \"top_p\": {\n              \"$omitEmpty\": {\n                \"$ref\": \"request.providerOptions.openai-response.top_p\"\n              }\n            },\n            \"max_output_tokens\": {\n              \"$omitEmpty\": {\n                \"$coalesce\": [\n                  {\n                    \"$ref\": \"request.extra.max_output_tokens\"\n                  },\n                  {\n                    \"$ref\": \"request.providerOptions.openai-response.max_output_tokens\"\n                  }\n                ]\n              }\n            },\n            \"tools\": {\n              \"$omitEmpty\": {\n                \"$ref\": \"request.providerOptions.openai-response.tools\"\n              }\n            },\n            \"tool_choice\": {\n              \"$omitEmpty\": {\n                \"$ref\": \"request.providerOptions.openai-response.tool_choice\"\n              }\n            },\n            \"text\": {\n              \"$omitEmpty\": {\n                \"$ref\": \"request.providerOptions.openai-response.text\"\n              }\n            },\n            \"reasoning\": {\n              \"$omitEmpty\": {\n                \"$ref\": \"request.providerOptions.openai-response.reasoning\"\n              }\n            },\n            \"previous_response_id\": {\n              \"$omitEmpty\": {\n                \"$ref\": \"request.providerOptions.openai-response.previous_response_id\"\n              }\n            },\n            \"store\": {\n              \"$omitEmpty\": {\n                \"$ref\": \"request.providerOptions.openai-response.store\"\n              }\n            },\n            \"metadata\": {\n              \"$omitEmpty\": {\n                \"$ref\": \"request.providerOptions.openai-response.metadata\"\n              }\n            },\n            \"stream\": {\n              \"$omitEmpty\": {\n                \"$ref\": \"request.providerOptions.openai-response.stream\"\n              }\n            },\n            \"truncation\": {\n              \"$omitEmpty\": {\n                \"$ref\": \"request.providerOptions.openai-response.truncation\"\n              }\n            },\n            \"user\": {\n              \"$omitEmpty\": {\n                \"$ref\": \"request.providerOptions.openai-response.user\"\n              }\n            }\n          }\n        },\n        \"agent\": {\n          \"method\": \"POST\",\n          \"path\": \"/responses\",\n          \"contentType\": \"application/json\",\n          \"body\": {\n            \"$merge\": [\n              {\n                \"$ref\": \"request.extra.agent.responses\"\n              },\n              {\n                \"model\": {\n                  \"$ref\": \"request.model\"\n                }\n              }\n            ]\n          }\n        },\n        \"response\": {\n          \"status\": \"succeeded\",\n          \"textPaths\": [\n            \"output_text\"\n          ],\n          \"reasoningPaths\": [\n            \"reasoning.summary.0.text\"\n          ],\n          \"usage\": {\n            \"$ref\": \"response.usage\"\n          },\n          \"errorPaths\": [\n            \"error.code\"\n          ],\n          \"messagePaths\": [\n            \"error.message\"\n          ]\n        },\n        \"agentResponse\": {\n          \"textPaths\": [\n            \"output_text\"\n          ],\n          \"reasoningPaths\": [\n            \"reasoning.summary.0.text\"\n          ],\n          \"toolCallsPath\": \"output\",\n          \"toolCallIdPaths\": [\n            \"call_id\",\n            \"id\"\n          ],\n          \"toolCallNamePaths\": [\n            \"name\"\n          ],\n          \"toolCallArgumentsPaths\": [\n            \"arguments\"\n          ]\n        }\n      }\n    ]\n  }\n}\n```\n<!-- YINGCE_MANIFEST_CONTRACT_END -->\n",
  "permissions": [
    "generation.run",
    "media.read"
  ],
  "configuration": {
    "fields": [
      {
        "name": "apiKey",
        "type": "secret",
        "label": "API Key",
        "required": true
      }
    ]
  },
  "contributes": {
    "providers": [
      {
        "id": "openai-response",
        "label": "OpenAI Responses",
        "capabilities": [
          "text"
        ],
        "scopes": [
          "admin.system-channel",
          "user.custom-channel",
          "canvas",
          "creation",
          "agent"
        ],
        "baseUrl": "https://api.openai.com",
        "requiresPublicMediaUrls": false,
        "auth": {
          "type": "bearer",
          "field": "apiKey"
        },
        "parameters": [
          {
            "name": "model",
            "type": "string",
            "required": true,
            "mapping": "model",
            "description": "上游模型 ID。"
          },
          {
            "name": "messages",
            "type": "message[]",
            "required": true,
            "mapping": "provider message container",
            "description": "包含历史消息和当前用户输入。"
          },
          {
            "name": "instructions",
            "type": "string",
            "required": false,
            "mapping": "system/instructions",
            "description": "系统指令。"
          },
          {
            "name": "temperature",
            "type": "number",
            "required": false,
            "mapping": "temperature",
            "description": "采样温度。"
          },
          {
            "name": "top_p",
            "type": "number",
            "required": false,
            "mapping": "top_p",
            "description": "核采样参数。"
          },
          {
            "name": "max_tokens",
            "type": "integer",
            "required": false,
            "mapping": "max_tokens/max_output_tokens",
            "description": "最大输出 token。"
          },
          {
            "name": "tools",
            "type": "array",
            "required": false,
            "mapping": "tools/toolConfig",
            "description": "工具定义。"
          },
          {
            "name": "tool_choice",
            "type": "object|string",
            "required": false,
            "mapping": "tool_choice",
            "description": "工具选择策略。"
          },
          {
            "name": "response_format",
            "type": "object",
            "required": false,
            "mapping": "response_format/text",
            "description": "结构化输出配置。"
          },
          {
            "name": "stream",
            "type": "boolean",
            "required": false,
            "mapping": "stream",
            "description": "流式开关；后台任务当前以最终响应归一。"
          }
        ],
        "create": {
          "method": "POST",
          "path": "/responses",
          "contentType": "application/json",
          "body": {
            "model": {
              "$ref": "request.model"
            },
            "input": {
              "$filter": {
                "from": {
                  "$ref": "request.messages"
                },
                "as": "message",
                "where": {
                  "$ne": [
                    {
                      "$ref": "message.role"
                    },
                    "system"
                  ]
                }
              }
            },
            "instructions": {
              "$omitEmpty": {
                "$ref": "request.instructions"
              }
            },
            "temperature": {
              "$omitEmpty": {
                "$ref": "request.providerOptions.openai-response.temperature"
              }
            },
            "top_p": {
              "$omitEmpty": {
                "$ref": "request.providerOptions.openai-response.top_p"
              }
            },
            "max_output_tokens": {
              "$omitEmpty": {
                "$coalesce": [
                  {
                    "$ref": "request.extra.max_output_tokens"
                  },
                  {
                    "$ref": "request.providerOptions.openai-response.max_output_tokens"
                  }
                ]
              }
            },
            "tools": {
              "$omitEmpty": {
                "$ref": "request.providerOptions.openai-response.tools"
              }
            },
            "tool_choice": {
              "$omitEmpty": {
                "$ref": "request.providerOptions.openai-response.tool_choice"
              }
            },
            "text": {
              "$omitEmpty": {
                "$ref": "request.providerOptions.openai-response.text"
              }
            },
            "reasoning": {
              "$omitEmpty": {
                "$ref": "request.providerOptions.openai-response.reasoning"
              }
            },
            "previous_response_id": {
              "$omitEmpty": {
                "$ref": "request.providerOptions.openai-response.previous_response_id"
              }
            },
            "store": {
              "$omitEmpty": {
                "$ref": "request.providerOptions.openai-response.store"
              }
            },
            "metadata": {
              "$omitEmpty": {
                "$ref": "request.providerOptions.openai-response.metadata"
              }
            },
            "stream": {
              "$omitEmpty": {
                "$ref": "request.providerOptions.openai-response.stream"
              }
            },
            "truncation": {
              "$omitEmpty": {
                "$ref": "request.providerOptions.openai-response.truncation"
              }
            },
            "user": {
              "$omitEmpty": {
                "$ref": "request.providerOptions.openai-response.user"
              }
            }
          }
        },
        "agent": {
          "method": "POST",
          "path": "/responses",
          "contentType": "application/json",
          "body": {
            "$merge": [
              {
                "$ref": "request.extra.agent.responses"
              },
              {
                "model": {
                  "$ref": "request.model"
                }
              }
            ]
          }
        },
        "response": {
          "status": "succeeded",
          "textPaths": [
            "output_text"
          ],
          "reasoningPaths": [
            "reasoning.summary.0.text"
          ],
          "usage": {
            "$ref": "response.usage"
          },
          "errorPaths": [
            "error.code"
          ],
          "messagePaths": [
            "error.message"
          ]
        },
        "agentResponse": {
          "textPaths": [
            "output_text"
          ],
          "reasoningPaths": [
            "reasoning.summary.0.text"
          ],
          "toolCallsPath": "output",
          "toolCallIdPaths": [
            "call_id",
            "id"
          ],
          "toolCallNamePaths": [
            "name"
          ],
          "toolCallArgumentsPaths": [
            "arguments"
          ]
        }
      }
    ]
  }
}