VividKit Crystal clear AI coding
VI
Choose Your Setup Get Started with AgentKit Continue with ClaudeKit Migrate from ClaudeKit to AgentKit Clean cutover from ClaudeKit What Is AgentKit? How AgentKit Works AgentKit Kits, Skills, Agents & Hooks AgentKit Runtime Support AgentKit Desktop App AgentKit Configuration AgentKit CLI Quick Reference AgentKit Skills Cheatsheet AgentKit Workflows AgentKit Statusline Update AgentKit CLI & Kits AgentKit Helper Clean reinstall of AgentKit AgentKit Troubleshooting What is ClaudeKit? ClaudeKit CLI Guide Command Finder Commands Cheat Sheet Workflow Recipes CLI Quick Reference Claude Mechanics Permission Modes ClaudeKit Hooks Coexistence ClaudeKit x Codex CK → Other coding tools How ClaudeKit Works Inside ClaudeKit Finding Your Unknowns Session Recovery Fix from Logs UI/UX Guide Accounts Switcher IDE Config Remote Control Codex App System Diagramming UI Review Gate Promotions Support VividKit

Quick Ref / Command

Engineer Kit

/ak:deploy

Detect and run deploys

Auto-detect or select a deployment target, load the matching platform reference, check CLI/auth, deploy, verify the URL, and maintain docs/deployment.md.

01

Read docs first

02

Scan configs

03

Analyze project

04

Recommend target

Rule 01

Detect before asking; ask only when target remains unclear.

Rule 02

Choose the cheapest suitable platform first.

Rule 03

Load one platform reference, not every deployment guide.

Rule 04

Verify the live deployment URL before claiming success.

Rule 05

Escalate advanced infrastructure and troubleshooting to ak:devops.

Execution Map

Core lanes from input to output, aligned with the command's real execution path.

01

Start

Input, scope, route

  1. 1 Read docs first If docs/deployment.md exists, parse platform, URL, deploy command, env vars, custom domain, and rollback notes.
  2. 2 Scan configs Detect platform signals such as vercel.json, netlify.toml, wrangler, fly.toml, railway, render, Procfile, tose, Coolify, Dokploy, GitHub Pages, GCP, AWS, or Digital Ocean files.
02

Work

Agent / skill execution

  1. 3 Analyze project Classify static site, SPA, SSR/full-stack, Node API, Python API, Docker app, or monorepo.
  2. 4 Recommend target When no target is detected, ask with up to four cost-optimized recommendations ordered by free tier or expected cost.
03

Verify

Gate, review, validation

  1. 5 Load platform ref Progressively disclose only the selected platform reference, never all platform files.
  2. 6 Check before publish Check CLI installation, authentication, .env handling, .gitignore, and platform-specific requirements before publishing.
04

Close

Report, handoff, artifact

  1. 7 Execute and verify Run the selected deploy command, read output, verify the deployment URL, and troubleshoot common failures.
  2. 8 Document result Create or update docs/deployment.md with platform, production URL, deploy command, env vars list, custom domain, rollback, and troubleshooting notes.

Syntax, arguments, subcommands

How to invoke the skill: syntax, positional arguments, shared options, then each subcommand with its own syntax and outcome.

Syntax

/ak:deploy [platform] [environment]

Arguments

[platform]

Hosting platform

Optional supported host such as Vercel, Netlify, Cloudflare, Railway, Fly.io, Render, Heroku, TOSE.sh, GitHub Pages, Coolify, Dokploy, GCP, AWS, DigitalOcean, or Vultr. Omit it to let the skill detect a host or ask you to choose. This does not pick the provider account, approve billing, or approve resource creation.

/ak:deploy cloudflare
[environment]

Environment label

Optional label such as production or preview. It is carried into target selection but does not create a standard default environment.

/ak:deploy cloudflare production

Modes / Routes

Vercel / Netlify / Cloudflare

Frontend and SSR

Research:
Config + framework scan
Review:
Env and build output
Validate:
Verify production URL
Railway / Render / Fly.io / TOSE.sh

Backend and full-stack

Research:
Runtime + Docker scan
Review:
Auth and service health
Validate:
Verify endpoint URL
GitHub Pages / Cloudflare Pages

Static site

Research:
Static asset scan
Review:
Build directory and routing
Validate:
Verify published pages
Coolify / Dokploy

Self-hosted

Research:
Docker/Compose signals
Review:
Server boundary
Validate:
Verify deployed service

Sample Prompt

Concrete invocations for each flag, subcommand, mode, or route available in the live workflow.

Auto-detect target Recommended 
/ak:deploy
Use when:
The goal is hosting or publishing the app, and docs or platform config may already identify the target.
Expected:
Reads docs/deployment.md first, scans supported config signals, loads only the matched platform reference, checks CLI/auth, deploys, verifies the URL, and updates deployment docs.
Explicit platform 
/ak:deploy cloudflare
Use when:
You already know the supported hosting platform but want the workflow to carry out its documented deploy path.
Expected:
Loads the Cloudflare reference only, checks the required CLI and authentication, runs the selected deploy command, verifies the resulting URL, and records operations in docs/deployment.md.
Platform and environment 
/ak:deploy vercel production
Use when:
The target platform and environment label are explicit before a production or preview publish.
Expected:
Carries the explicit environment into target selection, uses the Vercel platform reference, performs CLI/auth preflight, deploys, verifies the URL, and updates rollback notes.
First deploy choice 
/ak:deploy
Use when:
No deployment target is detected and the skill should recommend a supported host for the current project type.
Expected:
Analyzes project type, asks with up to four cost-ordered options including free-tier context, then continues only after a selected platform is available.

Handled Scope

  • Read docs first
  • Scan configs
  • Analyze project
  • Recommend target
  • Load platform ref
  • Check before publish

Artifacts Produced

Deployment document

Pattern

docs/deployment.md with platform, URL, command, environment variables, custom domain, rollback, and troubleshooting

Created after first successful deploy and updated when config changes.

Related skills

Next