Shengwang Integration

Workflow

Step 0: Ensure doc index exists (MANDATORY)

⚠️ 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 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.

Step 1: Collect kickoff information

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.

Step 2: Start with local references

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/.

Purpose	Route to
New request, vague, or missing details	intake
Credentials, AppID, REST auth	general
Download SDK, sample project, Token Builder, GitHub repo	Route to the relevant product module
Generate Token, token server, AccessToken2, RTC/RTM auth	token-server
ConvoAI operation (with details already known)	conversational-ai for SDK/sample-first guidance; use REST docs only as fallback reference
RTC SDK integration	rtc
RTM messaging / signaling	rtm
Cloud Recording	cloud-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.

Step 3: Research with doc fetching

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:

Local references in this skill
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
Fetched docs via the doc-fetching workflow
Fallback web search only if needed after doc fetching

Once Step 3 provides enough information, proceed with implementation.

Download Rules

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

voice-ai-integration