Skip to main content Home Creators modu-ai moai-adk moai-platform-deployment
moai-platform-deployment Deployment and hosting platform specialist covering Vercel, Railway, and Convex. Use when deploying applications, configuring edge functions, setting up continuous deployment, or managing serverless infrastructure.
Jump to install Skills Marketplace Discover and explore AI skills built by the community.
Install with Codex or Claude Copy this prompt, paste it into Codex, Claude, or another assistant, and let it review the skill page and install it for you.
Show prompt details A direct command skips the review prompt. Inspect the source before running it.
npx skills add https://github.com/modu-ai/moai-adk --skill moai-platform-deploymentThe command stays on one line. Scroll horizontally to inspect it before copying.
Prefer a local copy? Download the files currently available to SkillsMP.
Download Zip Downloading... More from this repository Harness learning subsystem coordinator. Produces Tier 4 auto-update proposal payloads consumed by the orchestrator (which surfaces them via AskUserQuestion) and orchestrates Apply/Rollback flows. Triggers when harness learning proposals are pending or learning lifecycle management is needed.
DEPRECATED — legacy 7-Phase meta-harness. Redirects to the v4 harness Builder (/moai:harness <natural-language request>) which replaces the static 7-Phase workflow with an orchestrator-direct 4-phase Builder (ANALYZE / PLAN / GENERATE / ACTIVATE) + a manifest-driven dynamic-workflow Runner. Retained as the redirect source for backward-compat invocation paths; the 7-Phase body below is preserved as historical reference, NOT for new harness creation.
MoAI unified orchestrator for autonomous development. Routes natural language or subcommands (plan, run, sync, project, fix, loop, mx, feedback, review, clean, codemaps, gate, e2e, harness, goal, todo) to specialized agents.
Related occupations SOC
Based on SOC occupation classification
name moai-platform-deployment description Deployment and hosting platform specialist covering Vercel, Railway, and Convex. Use when deploying applications, configuring edge functions, setting up continuous deployment, or managing serverless infrastructure.
license Apache-2.0 user-invocable false allowed-tools Read, Write, Edit, Grep, Glob, Bash(npm:*), Bash(npx:*), Bash(docker:*), Bash(git:*), WebFetch, mcp__context7__resolve-library-id, mcp__context7__get-library-docs metadata {"version":"2.0.0","category":"platform","status":"active","updated":"2026-02-09","platforms":"Vercel, Railway, Convex","tags":"deployment, hosting, vercel, railway, convex, edge, containers, serverless, real-time"} progressive_disclosure {"enabled":true,"level1_tokens":100,"level2_tokens":4500} triggers {"keywords":["deploy","deployment","hosting","vercel","railway","convex","edge functions","containers","docker","serverless","real-time","preview deployment","continuous deployment"],"agents":["expert-devops","expert-backend","expert-frontend"],"phases":["run","sync"]}
Deployment Platform Specialist
Comprehensive deployment platform guide covering Vercel (edge-first), Railway (container-first), and Convex (real-time backend).
Quick Platform Selection
When to Use Each Platform
Vercel - Edge-First Deployment:
Next.js applications with SSR/SSG
Global CDN distribution required
Sub-50ms edge latency critical
Preview deployments for team collaboration
Managed storage needs (KV, Blob, Postgres)
Railway - Container-First Deployment:
Full-stack containerized applications
Custom runtime environments
Multi-service architectures
Persistent volume storage
WebSocket/gRPC long-lived connections
Convex - Real-Time Backend:
Collaborative real-time applications
Reactive data synchronization
TypeScript-first backend needs
Optimistic UI updates
Document-oriented data models
Decision Guide
By Application Type
Web Applications (Frontend + API) :
Next.js → Vercel (optimal integration)
React/Vue with custom API → Railway (flexible)
Real-time collaborative → Convex + Vercel
Mobile Backends :
REST/GraphQL → Railway (stable connections)
Real-time sync → Convex (reactive queries)
Edge API → Vercel (global latency)
Full-Stack Monoliths :
Containerized → Railway (Docker support)
Serverless → Vercel (Next.js API routes)
Real-time → Convex (built-in reactivity)
By Infrastructure Needs
Compute Requirements :
Edge compute → Vercel (30+ edge locations)
Custom runtimes → Railway (Docker flexibility)
Serverless TypeScript → Convex (managed runtime)
Storage Requirements :
Redis/KV → Vercel KV or Railway
PostgreSQL → Vercel Postgres or Railway
File storage → Vercel Blob or Railway volumes
Document DB → Convex (built-in)
Networking Requirements :
CDN distribution → Vercel (built-in)
Private networking → Railway (service mesh)
Real-time WebSocket → Convex (built-in) or Railway
Common Deployment Patterns
Pattern 1: Next.js with Database Stack : Vercel + Vercel Postgres/KV
Deploy Next.js app to Vercel
Provision Vercel Postgres for database
Use Vercel KV for session/cache
Configure environment variables
Enable ISR for dynamic content
Best For : Web apps with standard database needs, e-commerce, content sites
Pattern 2: Containerized Multi-Service
Create multi-stage Dockerfile
Configure railway.toml for services
Set up private networking
Configure persistent volumes
Enable auto-scaling
Best For : Microservices, complex backends, custom tech stacks
Pattern 3: Real-Time Collaborative App Stack : Convex + Vercel/Railway (frontend)
Initialize Convex backend
Define schema and server functions
Deploy frontend to Vercel/Railway
Configure Convex provider
Implement optimistic updates
Best For : Collaborative tools, live dashboards, chat applications
Pattern 4: Hybrid Edge + Container Stack : Vercel (frontend/edge) + Railway (backend services)
Deploy Next.js frontend to Vercel
Deploy backend services to Railway
Configure CORS and API endpoints
Set up edge middleware for routing
Use private networking for Railway
Best For : High-performance apps, global distribution with complex backends
Pattern 5: Serverless Full-Stack Stack : Vercel (frontend + API routes) + Convex (backend)
Build Next.js app with API routes
Initialize Convex for data layer
Configure authentication (Clerk/Auth0)
Deploy frontend to Vercel
Connect Convex client
Best For : Rapid prototyping, startups, real-time web apps
Essential Configuration
Vercel Quick Start {
"$schema" : "https://openapi.vercel.sh/vercel.json" ,
"framework" : "nextjs" ,
"regions" : [ "iad1" , "sfo1" , "fra1" ] ,
"functions" : {
"app/api/**/*.ts" : {
"memory" : 1024 ,
"maxDuration" : 10
}
}
}
export const runtime = "edge"
export const preferredRegion = ["iad1" , "sfo1" ]
export async function GET (request : Request const country = request.geo ?.country || "Unknown"
return Response .json ({ country })
}
Railway Quick Start [build]
builder = "DOCKERFILE"
dockerfilePath = "Dockerfile"
[deploy]
healthcheckPath = "/health"
healthcheckTimeout = 100
restartPolicyType = "ON_FAILURE"
numReplicas = 2
[deploy.resources]
memory = "2GB"
cpu = "2.0"
# Builder stage
FROM node:20-alpine AS builder
WORKDIR /app
COPY package*.json ./
RUN npm ci
COPY . .
RUN npm run build
# Runner stage
FROM node:20-alpine
WORKDIR /app
ENV NODE_ENV=production
RUN addgroup -g 1001 -S nodejs && adduser -S appuser -u 1001
COPY --from=builder /app/node_modules ./node_modules
COPY --from=builder /app/dist ./dist
USER appuser
EXPOSE 3000
CMD ["node", "dist/main.js"]
Convex Quick Start import { defineSchema, defineTable } from "convex/server"
import { v } from "convex/values"
export default defineSchema ({
messages : defineTable ({
text : v.string (),
userId : v.id ("users" ),
timestamp : v.number (),
})
.index ("by_timestamp" , ["timestamp" ])
.searchIndex ("search_text" , {
searchField : "text" ,
filterFields : ["userId" ],
}),
})
import { useQuery, useMutation } from "convex/react"
import { api } from "../convex/_generated/api"
export function Messages (const messages = useQuery (api.messages .list )
const sendMessage = useMutation (api.messages .send )
if (!messages) return <div > Loading...</div > return (
<div >
{messages.map((msg) => (
<div key ={msg._id} > {msg.text}</div >
))}
</div >
CI/CD Integration
GitHub Actions - Vercel name: Deploy to Vercel
on:
push:
branches: [main ]
jobs:
deploy:
runs-on: ubuntu-latest
steps:
- uses: actions/checkout@v3
- uses: amondnet/vercel-action@v25
with:
vercel-token: ${{ secrets.VERCEL_TOKEN }}
vercel-org-id: ${{ secrets.ORG_ID }}
vercel-project-id: ${{ secrets.PROJECT_ID }}
GitHub Actions - Railway name: Deploy to Railway
on:
push:
branches: [main ]
jobs:
deploy:
runs-on: ubuntu-latest
steps:
- uses: actions/checkout@v3
- run: npm install -g @railway/cli
- run: railway up --detach
env:
RAILWAY_TOKEN: ${{ secrets.RAILWAY_TOKEN }}
GitHub Actions - Convex name: Deploy to Convex
on:
push:
branches: [main ]
jobs:
deploy:
runs-on: ubuntu-latest
steps:
- uses: actions/checkout@v3
- uses: actions/setup-node@v3
- run: npm ci
- run: npx convex deploy
env:
CONVEX_DEPLOY_KEY: ${{ secrets.CONVEX_DEPLOY_KEY }}
Advanced Patterns
Blue-Green Deployment (Vercel) Deploy new version, test on preview URL, then switch production alias using Vercel SDK for zero-downtime releases.
Multi-Region (Railway) Configure deployment regions in railway.toml:
[deploy.regions]
name = "us-west"
replicas = 2
[[deploy.regions]]
name = "eu-central"
replicas = 1
Optimistic Updates (Convex) const sendMessage = useMutation (api.messages .send )
const handleSend = (text : string sendMessage ({ text })
.then (() => console .log ("Sent" ))
.catch (() => console .log ("Failed, rolled back" ))
}
Platform-Specific Details For detailed platform-specific patterns, configuration options, and advanced use cases, see:
reference/vercel.md - Edge Functions, ISR, Analytics, Storagereference/railway.md - Docker, Multi-Service, Volumes, Scalingreference/convex.md - Reactive Queries, Server Functions, File Storagereference/comparison.md - Feature Matrix, Pricing, Migration Guides
Works Well With
moai-domain-backend for backend architecture patterns
moai-domain-frontend for frontend integration
.claude/rules/moai/languages/typescript.md for TypeScript best practices (auto-loaded via paths frontmatter).claude/rules/moai/languages/python.md for Python deployment on Railway (auto-loaded via paths frontmatter)moai-platform-auth for authentication integration
moai-platform-database for database patterns
Status: Production Ready
Version: 2.0.0
Updated: 2026-02-09
Platforms: Vercel, Railway, Convex
Common Rationalizations Rationalization Reality "I will configure the deployment platform after development is complete" Deployment configuration affects build output, environment variables, and runtime behavior. Configure early. "Preview deployments are optional" Preview deployments catch deployment-specific bugs before production. They cost little and save a lot. "Environment variables are the same across all environments" Production, staging, and development need different database URLs, API keys, and feature flags. One set of env vars is a security risk. "Serverless cold starts are negligible" Cold starts add 200-2000ms latency on first request. For user-facing APIs, this matters. Measure and mitigate. "I do not need a rollback strategy, I can just redeploy" Redeployment takes minutes. Rollback takes seconds. When production is down, seconds matter.
Red Flags
Production environment variables visible in build logs
No preview deployment configured for pull requests
Single environment used for both staging and production
No rollback mechanism documented or tested
Build artifacts include development dependencies or source maps
Verification
Refactor Notes R4 audit verdict (2026-04-23): REFACTOR — shrink triplet to Vercel-only primary; Railway/Convex as documentation-only
SPEC : SPEC-V3R2-WF-001 §6.2 line 271
Refactor scope (deferred to future sub-SPEC):
Elevate Vercel as primary platform with full guidance; demote Railway and Convex to doc-only references
Move Railway and Convex deep-dives into Level-3 modules
Add CI/CD pipeline patterns section (GitHub Actions, Vercel CI integration)
This skill is retained in v3.0 but its body will be restructured in a follow-up SPEC.
