HTML5 Studio

Free Course: the AI Picks Up Jira Tasks on Its Own

Jira is a task board — like a to-do list for a team. In one evening you go from "downloaded a folder" to "I send the AI a ticket number — it does the task and moves it across the board itself". Five steps in strict order, two prompts, nothing to configure by hand.

A task by its number

You type "pick up KAN-12" — the AI finds the task in Jira itself, reads the description, writes the code and moves the card across the board on its own: To Do → In Progress → Done. The board updates without you.

Everything is assembled for you

You download a ready-made project: a small web app, the rules for the AI and a placeholder file for the keys are already inside. Nothing to piece together — just add a key and type the first command.

Set up once — works forever

The rules are written into a CLAUDE.md file — the program's memory. Close your laptop, come back tomorrow, start a new chat — the AI remembers everything and works the same way.

Five steps — in strict order

1

Download the ready-made project

Open the repository via the button below → the green Code button → Download ZIP (or git clone if you know how). Unpack it into a folder. Everything is already inside: a small web app, the rules for the AI and a placeholder file for the keys.

Get the project on GitHub
2

Get a Jira key

Jira is a free task board (up to 10 people). Sign up at atlassian.com and create a project with a regular Kanban board. Now the key: open id.atlassian.com → Security → Create API token. It is a long password the AI will use to enter your Jira. It is shown once — copy it right away.

3

Let the AI write the settings itself

Open the downloaded folder in Claude Code (if you do not have the app yet — one terminal command: npm install -g @anthropic-ai/claude-code, then type claude). Paste prompt #1 (below) into the chat — the AI creates the keys file itself, tells you what goes where and checks the connection to your board.

4

Create a task in Jira

Right on the board, create a task "Create a Hello World page" with the description: "Add a /hello page that says: Hello, world!". Remember its number — for example, KAN-1.

5

Tell the AI to take the task

Paste prompt #2 into the chat: "Pick up KAN-1". And watch: the card moves to In Progress, code appears in the project, the task closes with a comment — what was done and which files were touched. The AI did that, not you.

Two prompts for the AI

A prompt is just text: copy it and paste it into the chat with the AI. There are only two: the first is setup (once), the second is work (every day).

1 Prompt #1 — setup (step 3)

The AI creates the keys file itself, asks you to fill in the values and checks the connection to your board. It will not change anything — it only looks around.

prompt-1-setup.md
Set up my Jira access following the rules in CLAUDE.md. Copy .env.example to .env and tell me what goes where: the Jira address, the account email, the API token from id.atlassian.com and the project key. I will fill in the values myself — I will not paste the token into the chat, and you must never print it to the chat, logs or code. When I say "done" — verify the connection: show my Jira account name, the board columns and the task list. Do not change anything yet — this is only a check.

If the AI showed your Jira account name and the task list — the connection works, move on to step 4. A 401 error means "not allowed in": check the email and the token in the .env file.

2 Prompt #2 — work (step 5)

That is it, setup is over. From now on your daily "command" is one line with a ticket number. The AI does the rest by the rules in the downloaded folder.

prompt-2-work.md
Pick up KAN-1.

Replace KAN-1 with your own issue number. Tomorrow and a month from now it works the same: the rules are already in the project.

🎉 Congratulations!

Your AI just picked a task off the board, did it and reported back in Jira. Not a demo, not a video — it works for you, on your board. Create more tasks and keep sending it numbers.

What comes next. Right now it is one task at a time. Our signature task pipeline runs dozens in parallel — with waves, task isolation and the launcher where tasks lie as cards and start with a click. And plugging it in is now even easier for you: Jira, the key and the AI link are already set up — part of the work is done. See the manual →

This pipeline is not theory

With the full version of the pipeline we build CHATBOSS.PRO — a service where Telegram bots are assembled with a mouse from ready-made blocks. Every new feature travels the whole route on its own: from task decomposition to the rollout on the live site.

500+
users have already built their bots
Visit CHATBOSS.PRO → Ask the AI consultant about the pipeline

💬 The AI consultant itself runs on the CHATBOSS.PRO platform — ask it about the pipeline, timelines and pricing right in Telegram.

Honestly about the course

Completely free

Five steps, a ready-made project and both prompts — take them and use them.

Your keys stay safe

The AI puts the Jira key into a file on your computer and hides it from publishing. It never prints it to the chat.

Everything is in the open

The prompts and the script are on GitHub: you can see exactly what you are running. Hit Watch — you will hear about new chapters.

Anyone can do it

The AI types the terminal commands itself. Your part — paste text into the chat and a couple of keys into a file.

Frequently Asked Questions

Is the course really free?

Yes: no payment, no sign-up, no fine print. It is a simplified version of the same approach we use ourselves and set up for clients.

Is it safe to give the AI a Jira key?

The key lives in a file on your computer and is hidden from publishing — the AI simply reads it from there, and the rules explicitly forbid it from showing the key in the chat, logs or code. And at any moment you can revoke the key at id.atlassian.com with one button — access disappears instantly.

What do I need to make it all work?

Jira (the free plan for up to 10 people is enough), the Claude Code program and an Anthropic subscription. Nothing else: the project comes ready-made, the AI writes its own settings. The whole thing takes one evening.

What comes after the course?

The course is one task at a time. Next comes the system: many tasks at once, in waves, two reviewers and deploys that go to a staging site first. All of that is in the manual "An AI Pipeline With Your Own Hands" with the Task Pipeline launcher included (preorder — 9,900 ₽). Or skip the DIY — we can set it up turnkey.

Questions about the course? Ask us — we will help

Stuck on a lesson, or the AI produced something odd — message us on Telegram, we will sort it out

Contact us