vibe-reconfigure

name	vibe-reconfigure
description	Reconfigure an existing vibe server. Change server role (master ↔ orchestrator), swap or add a deployment platform (Dokploy/Coolify), or add/remove edge deployment (Wrangler). Safe to run on a live server — detects current state first and only applies the delta.
argument-hint	["optional flags"]

Reconfigure

You are reconfiguring an already-provisioned vibe server. Detect current state first, then ask what to change. Only apply the delta — do not reinstall things that are already correctly set up.

Args: {{args}}

Privacy Rule — Redact Sensitive Values by Default

Do not print these in chat, even when a command you just ran returned them: server IP addresses, API keys / tokens, SSH private keys or fingerprints, passwords, or cloud credentials. Keep such values in shell variables and reference the variable rather than echoing the raw value; redact by omission when reporting progress (e.g. "Reconfigured — the IP/token is not shown in chat, ask if you need it."). SSH public keys may be shown. If the user explicitly asks for a value, give it in that one response only.

Phase 1: Detect Current State

Check the server config first — it's the fastest and most reliable source of truth:

cat /etc/vibemd/server.json 2>/dev/null || echo "NO_CONFIG"

If config exists, use it directly. No further detection needed. Show the user:

Current config (/etc/vibemd/server.json):

• Role: [role]
• Platform: [platform]
• Provider: [provider]
• Edge (Wrangler): [edge]
• AI coding CLIs: [ai_clis]
• Workers: [workers.length]

If no config, fall back to runtime detection:

# Server role indicator — cloud CLI presence means orchestrator
which hcloud doctl aws 2>/dev/null && echo "CLOUD_CLI_FOUND" || echo "NO_CLOUD_CLI"

# Deployment platform
docker ps --format '{{.Names}}' 2>/dev/null | grep -iE 'dokploy|coolify' || echo "NO_PLATFORM_CONTAINERS"
which dokploy 2>/dev/null && echo "DOKPLOY_CLI" || true

# AI coding CLIs
which claude codex 2>/dev/null || echo "NO_AI_CLIS"

# Edge
which wrangler 2>/dev/null && echo "WRANGLER_INSTALLED" || echo "NO_WRANGLER"

# OS / distro
uname -a && cat /etc/os-release 2>/dev/null | head -5

Summarize what you find before asking questions.

Phase 2: Interview

Important: Use AskUserQuestion (one call per question) for all questions in this phase. Do NOT use markdown checkbox lists. Each question maps to one AskUserQuestion call with up to 4 options (a built-in "Other" option is always available). Only show options that represent a change from the current state.

Question 1 — Server role:

Use AskUserQuestion (single-select) with the question: "Server role (current: [master / orchestrator]) — what would you like to do?"

Options:

1. Keep as-is
2. Switch to master/standalone (removes cloud CLI; machine manages only itself)
3. Switch to orchestrator (installs cloud CLI to provision and connect worker servers)

Question 2 — Deployment platform:

Use AskUserQuestion (single-select) with the question: "Deployment platform (current: [Dokploy / Coolify / both]) — what would you like to do?"

Options:

1. Keep as-is
2. Switch to Dokploy (lightweight, Docker-native)
3. Switch to Coolify (more features, great UI)
4. Add the other platform alongside current one (not recommended — only for evaluating both with separate apps)

Question 3 — Edge deployment (Wrangler):

Use AskUserQuestion (single-select) with the question: "Edge deployment / Wrangler (current: [installed / not installed]) — what would you like to do?"

Options:

1. Keep as-is
2. Install Wrangler (Cloudflare Workers, Pages, DNS)
3. Remove Wrangler

Question 4 — AI coding CLIs (part A):

Use AskUserQuestion (single-select) with the question: "AI coding CLIs (current: [Claude Code / Codex / both / none]) — keep as-is or make a change?"

Options:

1. Keep as-is
2. Make a change (answer the next question)

If the user chooses "Make a change", ask a follow-up:

Question 4 — AI coding CLIs (part B):

Use AskUserQuestion (single-select) with the question: "Which AI coding CLI setup do you want?"

Options:

1. Install Claude Code only (@anthropic-ai/claude-code)
2. Install Codex CLI only (@openai/codex)
3. Install both Claude Code and Codex CLI
4. Remove all AI coding CLIs

Wait for all answers before making any changes.

Phase 3: Apply Changes

Only apply what changed. Skip anything marked "Keep as-is."

Role: Switch to Orchestrator

Install the cloud CLI matching the provider originally used to provision the server. See ../vibemd/references/orchestrator-setup.md.

Role: Switch to Master

Remove the cloud CLI:

# Hetzner
apt-get remove -y hcloud-cli 2>/dev/null || true
rm -f /usr/local/bin/hcloud

# DigitalOcean
rm -f /usr/local/bin/doctl

# AWS
pip3 uninstall -y awscli 2>/dev/null || rm -f /usr/local/bin/aws

Confirm removal: which hcloud doctl aws 2>/dev/null || echo "Cloud CLI removed"

Platform: Switch or Add

Follow the platform installation steps in ../vibemd/references/server-setup.md (Step 4). If switching away from an existing platform, warn the user first:

Warning: Switching platforms will not automatically migrate your existing apps, databases, or volumes. Make sure everything is backed up before proceeding.

Only proceed after explicit confirmation.

Edge: Install Wrangler

bun add -g wrangler
wrangler --version

wrangler login opens a browser OAuth flow and will not work on a headless server. Use the API token approach instead (create a token at Cloudflare dashboard → My Profile → API Tokens, with Workers/Pages/DNS permissions):

export CLOUDFLARE_API_TOKEN=<your-cloudflare-api-token>
echo 'export CLOUDFLARE_API_TOKEN=<your-cloudflare-api-token>' >> ~/.bashrc
wrangler whoami   # verify auth

Edge: Remove Wrangler

bun remove -g wrangler

Phase 4: Update Server Config

After applying all changes, update /etc/vibemd/server.json to reflect the new state. If the file doesn't exist yet, create it.

# Read existing config or start from scratch
CONFIG=$(cat /etc/vibemd/server.json 2>/dev/null || echo '{"workers":[],"provisioned_at":"unknown"}')

mkdir -p /etc/vibemd
echo "$CONFIG" | jq \
  --arg role "<new-role>" \
  --arg platform "<new-platform>" \
  --arg edge "<true|false>" \
  --argjson ai_clis '["<claude-code|codex>"]' \
  '.role = $role | .platform = $platform | .edge = ($edge == "true") | .ai_clis = $ai_clis | .reconfigured_at = now | todate' \
  > /tmp/vibemd-server.json && mv /tmp/vibemd-server.json /etc/vibemd/server.json

echo "Updated /etc/vibemd/server.json:"
cat /etc/vibemd/server.json

Only update the fields that changed — leave workers, provider, provisioned_at, and other unchanged fields intact.

Completion Checklist

• Current state read from config (or detected from runtime)
• User confirmed desired changes
• Server role updated (if changed)
• Deployment platform updated (if changed)
• Wrangler installed or removed (if changed)
• AI coding CLIs installed or removed (if changed)
• /etc/vibemd/server.json updated
• No unintended side effects (existing apps still running)