Deploy logo

Deploy

Community
notque
deploy

Deploy and infrastructure: web deploy, cron, package install, shell configuration.

Overview

Publishernotque
Repositoryvexjoy-agent
Skill namedeploy
Stars
443
Forks
50
Bundled files
23
LicenseMIT
Links
  • Markdown instructions

    A SKILL.md file the model loads on demand, so it only costs tokens when a request actually matches.

  • Works with any LLM

    AI skills are plain Markdown, not provider-specific code, so this works with GPT, Claude, Gemini, Grok, or a local model.

  • 23 bundled files

    Scripts, templates, and references the model can read while it works. Files are read-only and never executed.

  • Open source

    Published by notque on GitHub. Read the source before you install it.

Installation

Install the Deploy AI skill in TypingMind to use it with any LLM, or drop it into another agent that reads SKILL.md.

1

Install in TypingMind

TypingMind installs a skill straight from its GitHub folder — it reads SKILL.md, bundles the resource files, and stores the result locally.

  1. Open the app and go to Plugins → Skills.
  2. Choose "Install from GitHub".
  3. Paste the skill folder URL below and confirm.
    https://github.com/notque/vexjoy-agent/tree/main/skills/infrastructure/deploy
  4. Enable the skill in any chat where you want it available.
Plugins → Skills → Add skill → From GitHub URL, then paste the folder URL and press Continue.
2

Install in another agent

Any agent that reads the Agent Skills format can use this skill — copy the folder into that agent's skills directory.

Claude Code — .claude/skills
git clone --depth 1 https://github.com/notque/vexjoy-agent.git /tmp/vexjoy-agent
mkdir -p .claude/skills
cp -r /tmp/vexjoy-agent/skills/infrastructure/deploy .claude/skills/deploy
Restart Claude Code after copying so it picks up the new skill.

Use it in TypingMind

Enable Deploy in any TypingMind chat and the model takes it from there. Its name and description sit in the system prompt, and the moment a request matches, the model loads the full instructions itself — you never invoke it by hand, and it costs no tokens until it is actually used.

The model loads Deploy on its own as soon as a request matches it.

Works with any AI model

AI skills are plain Markdown instructions rather than provider-specific code, so Deploy is not tied to the model it was written for. Install it once in TypingMind and use it with GPT-5, Claude, Gemini, Grok, DeepSeek, Mistral, Llama, or a local model you run yourself — all on your own API keys.

  • Loaded only when it is needed

    The system prompt carries just the name and description. The instructions are fetched on the first matching request, so an idle skill costs nothing.

  • Switch models mid-chat

    Because the skill is instructions rather than code, changing model does not break it — the next model reads the same SKILL.md.

Skill instructions

This is the SKILL.md content the model loads. Read it before installing — a skill is instructions your model will follow.

Deploy Skill

Six modes. Match the request to a section.

SignalSection
Deploy site, HTTPS, nginx, go live, public siteA. Public Web Deploy
Cron job, scheduled task, audit cron scriptC. Cron Automation
Headless agent, wrapper script, recurring Claude taskD. Headless Cron Creator
Install toolkit, verify setup, health checkE. Toolkit Install
Fish shell, Zsh, shell config, migrationF. Shell Configuration
Background process, nohup, trap, PID, signalsG. Process Management

When the request spans modes, compose the relevant sections.


A. Public Web Deploy

Serve public sites through nginx/Caddy/Apache -- never a raw dev server. Local preview binds 127.0.0.1; public sites require HTTPS + hardened nginx.

5-phase workflow: DNS -> Web Server -> HTTPS -> Hardening -> Verify.

  1. DNS: Create A/AAAA record. Verify: dig +short A <fqdn>. Gate: resolves to intended IP.
  2. Web Server: nginx server block with explicit docroot. Backend proxied to 127.0.0.1. Test: nginx -t && systemctl reload nginx.
  3. HTTPS: certbot --nginx -d <fqdn>. Verify renewal: certbot renew --dry-run. Verify redirect: curl -sI http://<fqdn> | grep -iE '301|308'.
  4. Hardening: Firewall (ufw allow 80/443/SSH only), nginx deny rules (location ~ /\. { deny all; }), security headers (HSTS, X-Content-Type-Options, CSP in Report-Only), rate limiting (limit_req), fail2ban.
  5. Verify: Run the 13-item security checklist. Spot-check: curl -s https://<fqdn>/.env -w '%{http_code}\n' (expect 403/404), ss -tlnp (no app ports public).

Load references/public-web-deploy.md for the full 13-item checklist, nginx config blocks, and error handling.


C. Cron Automation

Static analysis of cron scripts against a 9-point reliability checklist. Read-only -- never execute scripts.

  1. DISCOVER: Locate scripts in scripts/*.sh, cron/*.sh, jobs/*.sh. Verify shell shebang.
  2. AUDIT: Run all 9 checks via regex (verify matches not in comments):
#CheckSeverityKey patterns
1Error handlingCRITICALset -e, set -o errexit
2Exit code checkingHIGH$?, if [ $? -eq
3Logging with timestampsHIGH>> *.log, $(date)
4Log rotationMEDIUMfind -mtime -delete, logrotate
5Working directoryHIGHcd "$(dirname", SCRIPT_DIR=
6PATH environmentMEDIUMPATH=, export PATH
7Lock file / concurrencyHIGH.lock, flock, .pid
8Cleanup on exitMEDIUMtrap ... EXIT
9Failure notificationLOWmail -s, curl *webhook
  1. REPORT: Per-script scores with paste-ready fixes for every FAIL/WARN. Aggregate summary for multi-script audits.

Load references/cron-automation.md for the best-practices reference script.


D. Headless Cron Creator

Create headless Claude Code cron jobs. All crontab mutations go through crontab-manager.py.

  1. PARSE: Extract name (kebab-case), prompt, schedule, workdir, budget ($2.00 default).
  2. GENERATE: python3 ~/.claude/scripts/crontab-manager.py generate-wrapper --name <name> .... Verify: flock, --permission-mode auto, --max-budget-usd, tee logging, dry-run default.
  3. VALIDATE: bash -n scripts/<name>-cron.sh. Run 9-point cron checklist.
  4. INSTALL: Dry-run first (--dry-run), ask user confirmation, then install.
  5. REPORT: Script path, schedule, log dir, budget, management commands.

Load references/headless-cron-creator.md for the full methodology and schedule conversion table.


E. Toolkit Install

  1. Run python3 ~/.claude/scripts/install-doctor.py check and python3 ~/.claude/scripts/toolkit-health.py --json.
  2. If issues: guide user to ./install.sh --symlink. Fix permissions and deps as needed.
  3. Show inventory: python3 ~/.claude/scripts/install-doctor.py inventory.
  4. Show MCP status: python3 ~/.claude/scripts/mcp-registry.py list.
  5. Orient: /do, /comprehensive-review, /install commands.

Load references/install.md for full diagnostic steps and error handling.


F. Shell Configuration

Detect target shell first. Fish and Zsh have incompatible syntax.

ConceptFishZsh
Variable assignmentset -gx VAR valueexport VAR=value
PATH managementfish_add_pathtypeset -U path; path=(...)
Completionscompletions/ directorycompinit + fpath
Conditionalstest, not [[ ]][[ ]] preferred
Interactive guardstatus is-interactive[[ -o interactive ]]

Load the shell-specific reference matching the task:

TaskFishZsh
Full configreferences/fish-shell-config.mdreferences/zsh-shell-config.md
Migration from Bashreferences/fish-bash-migration.mdreferences/zsh-bash-migration.md
Variables, special varsreferences/fish-quick-reference.mdreferences/zsh-quick-reference.md
Error audit, failure modesreferences/fish-preferred-patterns.mdreferences/zsh-preferred-patterns.md
Dev tool integrationreferences/fish-tool-integrations.mdreferences/zsh-tool-integrations.md

G. Process Management

Match the need to the pattern:

GoalCommand
Background job, exits with shellcmd &
Survive shell exitnohup cmd > log 2>&1 &
Detach from job controlcmd & disown
Full detach, new sessionsetsid cmd > log 2>&1 < /dev/null &
System daemonsystemd unit file

All backgrounded processes must redirect stdio: > out 2>&1 < /dev/null &. Order matters: > out 2>&1 is correct; 2>&1 > out is almost always a bug.

Key rule: verify the observable state (port free, PID dead) after every kill -- do not assume kill succeeded.

Load deep references for specific tasks:

TaskReference
Process lifecycle patternsreferences/shell-process-patterns.md, references/starting-processes.md
PID capture and resolutionreferences/pid-resolution.md
Signal and trap disciplinereferences/signals-and-traps.md
Kill verification, stale PIDreferences/cleanup-verification.md
Shell gotchas (set -e, `

Deep References

All references below are >100 lines of domain-specific content. Load as directed by the sections above.

ReferenceLinesDomain
references/public-web-deploy.md259Full public deploy workflow + security checklist
references/cron-automation.md1949-point cron audit checklist
references/headless-cron-creator.md150Headless cron job creation
references/install.md195Toolkit install and health check
references/shell-error-handling.md226Shell error handling patterns
references/concurrency-and-locks.md219Concurrency and locking
references/logging-and-rotation.md231Logging and rotation
references/starting-processes.md152Process start patterns
references/shell-process-patterns.md247Process lifecycle patterns
references/pid-resolution.md200PID capture and resolution
references/signals-and-traps.md259Signal and trap discipline
references/cleanup-verification.md253Kill-and-check verification
references/preferred-patterns.md354Shell gotchas and fixes
references/fish-shell-config.md247Fish shell configuration
references/fish-bash-migration.md149Bash-to-Fish migration
references/fish-quick-reference.md229Fish variable scope guide
references/fish-preferred-patterns.md240Fish failure modes
references/fish-tool-integrations.md318Fish dev tool patterns
references/zsh-shell-config.md310Zsh shell configuration
references/zsh-bash-migration.md248Bash-to-Zsh migration
references/zsh-quick-reference.md332Zsh parameter expansion
references/zsh-preferred-patterns.md290Zsh failure modes
references/zsh-tool-integrations.md368Zsh dev tool patterns

Bundled files

The model reads these on demand while the skill is loaded. They are exposed as readable files and are never executed.

Frequently asked questions

What does the Deploy AI skill do?

Deploy and infrastructure: web deploy, cron, package install, shell configuration.

Why use Deploy on TypingMind?

Because you install it once and use it with any model. Deploy is plain Markdown rather than provider-specific code, so the same skill runs on GPT-5, Claude, Gemini, Grok, or a local model — and you can switch model mid-chat without it breaking. TypingMind runs on your own API keys, so you pay providers directly instead of a per-seat subscription, and your skills and chats stay in your own storage.

How do I install Deploy in TypingMind?

Open Plugins → Skills → Install from GitHub in TypingMind and paste https://github.com/notque/vexjoy-agent/tree/main/skills/infrastructure/deploy. TypingMind reads its SKILL.md and bundles its files and installs it as a skill you can enable per chat.

Which AI models can use Deploy?

Any model you connect in TypingMind. AI skills are plain Markdown instructions rather than provider-specific code, so GPT, Claude, Gemini, Grok, and local models can all load this skill when a request matches it.

How many AI models can I use with Deploy?

As many as you like. As long as a model supports skills, you can use Deploy with it — GPT, Claude, Gemini, Grok, DeepSeek, Mistral, Llama and more — all on TypingMind with your own API keys.

Is the Deploy AI skill free?

Yes. It is published on GitHub by notque under the MIT license. You only pay your own AI provider for the tokens you use.

What are AI skills?

An AI skill is a reusable instruction bundle that teaches an AI model how to do one specific task. It follows the open Agent Skills format: a SKILL.md file with a name and description, plus any scripts, templates or reference files the model may need. The model reads the instructions only when your request matches the skill, so an installed skill costs nothing until it is used.

How are AI skills different from plugins or MCP servers?

A plugin or MCP server gives a model new tools to call — code that runs somewhere and returns a result. An AI skill gives the model knowledge and process instead: how to approach a task, which steps to follow, what good output looks like. Skills are plain Markdown, so they need no server, no API key and no runtime, and they work with any model.

View all

Set up your own AI workspace now

Get notified about new features and future giveaways by subscribing to our newsletter 👇