Back to Details

voice-ai-integration

Compare original and translation side by side

🇺🇸

Original

English
🇨🇳

Translation

Chinese

Shengwang Integration

声网集成

Workflow

工作流程

Step 0: Ensure doc index exists (MANDATORY)

步骤0：确保文档索引存在（强制要求）

⚠️ This step is NON-NEGOTIABLE. Execute it BEFORE any routing, intake, or code generation.
Check if 
references/docs.txt
exists. If not (or if this is a fresh project), download it immediately:
bash
bash skills/voice-ai-integration/scripts/fetch-docs.sh
This file is the documentation index — all doc lookups depend on it. Do NOT proceed to Step 1 until this file exists or the download has been attempted. If download fails, proceed with local reference docs and fallback URLs.
⚠️ 此步骤为硬性要求。在进行任何路由、信息收集或代码生成之前必须执行。
检查
references/docs.txt
是否存在。如果不存在（或者这是一个新项目），请立即下载：
bash
bash skills/voice-ai-integration/scripts/fetch-docs.sh
该文件是文档索引——所有文档查询都依赖于它。在该文件存在或已尝试下载之前，请勿进入步骤1。如果下载失败，请使用本地参考文档和备用URL继续。

Step 1: Collect kickoff information

步骤1：收集启动信息

Use intake to collect kickoff information. Ask only for details the user has not already provided.
Collect only the details needed to remove implementation blockers:
  • User's use case / target solution
  • Main Shengwang product
  • Platform or client stack
  • Backend language if relevant
  • Any key technical details already known that affect routing or implementation
Use a conversational flow:
  • Infer obvious context from the user's request when it is safe to do so
  • Ask only for missing details that block routing or implementation
  • Stop asking as soon as there is enough information to continue
ConvoAI has a special intake mode:
  • If ConvoAI is clearly the primary product, switch to the consolidated ConvoAI intake in intake/convoai.md
  • Ask for all unresolved kickoff fields plus unresolved ConvoAI provider/config fields in one message
  • Show numbered choices for each unresolved field and ask for a one-line numeric reply
  • Do not repeat fields the user already answered
  • For ConvoAI implementation, prefer the official sample repo, 
    agora-agent-server-sdk
    on the server side, and 
    agora-agent-client-toolkit
    on the client side when the target stack supports it, over building directly from the REST spec
  • Treat raw REST as a fallback only for unsupported operations, debugging, or when the user explicitly asks for a REST-first implementation
For ConvoAI, the user must still explicitly answer or confirm any unresolved 
Other
follow-up value before implementation.
For unresolved ConvoAI fields with defaults, keep them visible and treat omission as an explicit default confirmation. This includes 
Platform = Web
and 
Backend = Python
. If the first consolidated reply is incomplete, ask only a narrow follow-up for the unresolved mandatory blocker.
If the user already gave enough information, do not repeat questions. Produce a lightweight kickoff recap, then continue automatically unless a required detail is still missing.
使用intake收集启动信息。仅询问用户尚未提供的细节。
仅收集消除实现障碍所需的细节：
  • 用户的用例/目标解决方案
  • 主要使用的声网产品
  • 平台或客户端技术栈
  • 相关的后端语言（如有需要）
  • 已知的、会影响路由或实现的关键技术细节
采用对话式流程：
  • 在安全的前提下，从用户的请求中推断明显的上下文
  • 仅询问缺失的、会阻碍路由或实现的细节
  • 一旦获得足够的信息，立即停止询问
ConvoAI有特殊的信息收集模式：
  • 如果ConvoAI显然是主要产品，请切换到intake/convoai.md中的整合式ConvoAI信息收集流程
  • 在一条消息中询问所有未解决的启动字段以及未解决的ConvoAI提供商/配置字段
  • 为每个未解决的字段显示编号选项，并要求用户以单行数字回复
  • 不要重复用户已经回答过的字段
  • 对于ConvoAI实现，当目标技术栈支持时，优先使用官方示例仓库、服务端的
    agora-agent-server-sdk
    和客户端的
    agora-agent-client-toolkit
    ，而非直接基于REST规范构建
  • 仅在不支持的操作、调试或用户明确要求优先使用REST实现时，才将原生REST作为备选方案
对于ConvoAI，在开始实现之前，用户必须明确回答或确认任何未解决的「其他」后续值。
对于带有默认值的未解决ConvoAI字段，需保持可见，并将用户未提及视为明确确认默认值。这包括「平台 = Web」和「后端 = Python」。 如果第一次整合回复不完整，请仅针对未解决的强制阻碍因素进行窄范围的后续询问。
如果用户已经提供了足够的信息，请勿重复提问。生成简洁的启动信息摘要，然后自动继续，除非仍缺少必填细节。

Step 2: Start with local references

步骤2：从本地参考文档开始

Use the kickoff summary plus the route table below to select the correct local reference module. If the available information is sufficient, begin implementation using the existing local docs under 
references/
.
PurposeRoute to
New request, vague, or missing detailsintake
Credentials, AppID, REST authgeneral
Download SDK, sample project, Token Builder, GitHub repoRoute to the relevant product module
Generate Token, token server, AccessToken2, RTC/RTM authtoken-server
ConvoAI operation (with details already known)conversational-ai for SDK/sample-first guidance; use REST docs only as fallback reference
RTC SDK integrationrtc
RTM messaging / signalingrtm
Cloud Recordingcloud-recording
If Step 2 provides enough information for implementation, proceed. If essential information is still missing or the local references are not enough, continue to Step 3.
使用启动信息摘要加上下方的路由表选择正确的本地参考模块。如果可用信息足够，请使用
references/
下的现有本地文档开始实现。
用途路由至
新请求、模糊请求或缺少细节intake
凭证、AppID、REST认证general
下载SDK、示例项目、Token生成工具、GitHub仓库路由至相关产品模块
生成Token、Token服务器、AccessToken2、RTC/RTM认证token-server
ConvoAI操作（已知细节）conversational-ai（优先SDK/示例指导；仅在必要时将REST文档作为备选参考）
RTC SDK集成rtc
RTM消息/信令rtm
云端录制cloud-recording
如果步骤2提供了足够的实现信息，请继续。如果仍缺少关键信息或本地参考文档不足，请进入步骤3。

Step 3: Research with doc fetching

步骤3：通过文档获取进行调研

Use references/doc-fetching.md to fetch more comprehensive documentation. Do this only after Step 2, when the local references are insufficient for the requested implementation.
Research order:
  1. Local references in this skill
  2. For ConvoAI, inspect the matching sample repo, 
    agora-agent-server-sdk
    , and 
    agora-agent-client-toolkit
    path before using REST docs as a design source
  3. Fetched docs via the doc-fetching workflow
  4. Fallback web search only if needed after doc fetching
Once Step 3 provides enough information, proceed with implementation.
使用references/doc-fetching.md获取更全面的文档。仅在步骤2之后、本地参考文档不足以支持所需实现时执行此步骤。
调研顺序：
  1. 本技能中的本地参考文档
  2. 对于ConvoAI，在将REST文档作为设计来源之前，先检查匹配的示例仓库、
    agora-agent-server-sdk
    和
    agora-agent-client-toolkit
    路径
  3. 通过文档获取流程获取的文档
  4. 仅在文档获取后仍有需要时，才使用备选网络搜索
一旦步骤3提供了足够的信息，即可开始实现。

Download Rules

下载规则

  • Use 
    git clone --depth 1 <url>
    — GitHub URLs must be repo root only (no branch/subdirectory paths)
  • On any download failure: report the error, provide the URL for manual download, never silently skip
  • 使用
    git clone --depth 1 <url>
    — GitHub URL必须仅为仓库根目录（不包含分支/子目录路径）
  • 若下载失败：报告错误，提供手动下载的URL，切勿静默跳过

Links

链接