AI agents that actually deploy instead of just chatting

Most AI agents are just a chat box prompt that guesses your project structure, but Kiro Crew packages agents, skills, and cron jobs into a real deployable app. I built a Daily Standup Bot because remembering yesterday's work is the hardest part of my morning.

The bot pulls my git commits from the last 24 hours and turns them into a What I Did / What's Blocked / What's Next list. It runs every weekday at 9 AM and posts to a custom dashboard. No more staring at a blank Slack channel for ten minutes.

Here is the blueprint to build it yourself.

The Architecture

An app here is a bundle, not just a script. You mix agents, MCP servers, and UI pages. For this standup bot:

standup-bot/
├── app.json ← manifest (identity + resources)
├── agents/
│ └── standup-agent.json ← agent definition
├── skills/
│ └── standup-format/
│ └── SKILL.md ← formatting rules
└── ui/
 └── src/App.tsx ← dashboard page

Step-by-Step Deployment

1. The Manifest: Create app.json. This tells the system what your app is and what files to load.

{
 "name": "standup-bot",
 "version": "1.0.0",
 "displayName": "Daily Standup Bot",
 "description": "Auto-generates standup notes from git commits.",
 "author": "sarvar_04",
 "agents": ["agents/standup-agent.json"],
 "skills": ["skills/standup-format"],
 "ui": {
 "entry": "dist/index.mjs",
 "pages": [{
 "route": "/apps/standup-bot",
 "label": "Standups",
 "icon": "ClipboardList"
 }]
 },
 "crons": [{
 "name": "morning-standup",
 "cron_expr": "0 9 * * 1-5",
 "message": "Generate today's standup summary from my recent git logs."
 }]
}

1. The Agent: Create agents/standup-agent.json. This sets the LLM's persona. Keep it focused so it doesn't wax poetic about commit messages.

1. The Skill: Put a markdown file at skills/standup-format/SKILL.md. It teaches the agent how to format output so it isn't a wall of text.

1. The Dashboard: Write a React component in ui/src/App.tsx for a sidebar page. The bot posts summaries here.

1. The Automation: The crons section in the manifest handles scheduling. Skip system-level crontabs and let the orchestrator manage it.

The key advantage is isolation. You're not adding a prompt to a global list. You're building a versioned tool that others can install with one command. It's an App Store for LLM agents that cuts out the daily prompt repetition.