mcp server

xidao-cookbook

Migration guides, rollout checklists, cost-control patterns, and LLM routing recipes for OpenAI-compatible APIs. Covers OpenRouter alternatives, failover strategies, and production deployment.

XidaoApi/xidao-cookbookcommunitylistedunreviewed

Use tool

Set up

Runs on your computer. Pick your agent and operating system.

This project does not publish an install method we can turn into a config. Follow the README instead.

README

synced Oct 23, 2026

XiDao Cookbook

Practical migration guides and production patterns for teams using an OpenAI-compatible AI API.

Why this repo exists

Most AI integration content stops at the first successful request. Production teams usually need help with harder problems:

safe provider switching
phased rollout and regression testing
fallback and routing strategy
cost control by workload
observability when traffic spans multiple model options

This cookbook focuses on those higher-intent, production-facing problems first.

Sections

1. Migration and rollout

switch from OpenAI API with minimal code changes
verify which SDK assumptions your app depends on
run rollout checklists before moving real production traffic
separate output-quality risk from integration-risk during migration

2. Reliability and routing

fallback versus blind retry behavior
health-aware routing patterns
latency-tier and task-tier model selection
staged rollout patterns that reduce outage blast radius

3. Cost optimization

prompt trimming
model routing by task value
caching repeated requests
token and request visibility for cost analysis

4. Framework examples

Python OpenAI SDK
Node.js OpenAI SDK
LangChain-ready notes
Vercel AI SDK integration notes

Repo relationship map

xidao-python-examples → minimal Python usage examples
xidao-nodejs-examples → minimal Node.js usage examples
llm-failover-router-demo → code-first reliability and routing examples
xidao-cookbook → migration, rollout, routing, and cost guides