32dots HEIDELBERG AI
Curriculum › Getting Started › 01
Session 1

Your first AI agent

easy
USE 0 - 15 min

Run it — before we explain anything

Login: use the email you signed up with at curriculum.32dots.de/signup + password cos2026. No account? Use demo@cos.uni-heidelberg.de / cos2026.
https://n8n.32dots.de ↗
  1. 1 Open https://n8n.32dots.de in your browser. Log in with your own student email + password cos2026. (No demo? Sign up at https://curriculum.32dots.de/signup or use demo@cos.uni-heidelberg.de.)
  2. 2 Log in with the credentials shown above.
  3. 3 In the left sidebar, click Workflows. Find 'Session 01 — Your first AI agent'.
  4. 4 Click the ⋯ menu on the workflow → Duplicate. Edit your copy (not the shared original).
  5. 5 Open the workflow. Click the Chat button in the bottom-right corner.
  6. 6 Ask: 'What is CRISPR and how does it work?'
  7. 7 Ask a follow-up: 'What are the main ethical concerns about using CRISPR in humans?'
  8. 8 Ask something it probably cannot answer well: 'What did Professor Müller publish last week?'
✓

You got a response to all three questions. The third answer was vague or wrong — and that is the most important result of this exercise.

UNDERSTAND 15 - 60 min

Anatomy of an AI agent — every node explained

💬When chat message …TriggerTRIGGER🧠AI AgentOrchestratorORCHESTRATOR⚡Groq gpt-oss-120b …Language modelCHAT MODEL🧵Simple Memory (win…Short-term memoryMEMORY🌐HTTP Request ToolToolTOOL🪆AI Agent Tool (sub…DelegationTOOL

The solid arrows show the main flow; the dashed lines are how Model, Memory, and Tools plug into the Agent on its three input ports. This is the same shape n8n draws on its canvas.

Key concept

An AI agent is four things wired together: (1) a language model — the pattern engine that writes text but knows nothing of today's world, (2) a system prompt — the rules that shape its behaviour, (3) memory — what it remembers between turns, (4) tools — how it reaches out of its own head and into the real world. Change any of the four and you get a different agent. The model is the smallest part of the job — the plumbing around it is where the engineering happens.

  1. ?Open the AI Agent node. Read its system prompt. Can you identify the four parts (role, constraints, style, output format)? Which part is strongest, which is weakest?
  2. ?Open the Simple Memory node. How many turns does it keep? What would break if you set it to 1? What would become expensive if you set it to 100?
  3. ?The agent has three tools wired to it. How does the agent decide which one to call — and how does the tool's description influence that decision?
  4. ?Why would you split an agent with 10 tools into a main agent + 2 sub-agents instead of just giving it all 10 tools?
BUILD 60 - 90 min

Make it yours — prompt, tools, memory

Your task

Duplicate the shared workflow. Then (1) rewrite the system prompt using the four-part anatomy, (2) add or remove one tool, (3) run a two-turn chat that proves memory is working.

  1. 1 In n8n, open the workflow. Click ⋯ → Duplicate. Work in your copy only — never edit the shared original.
  2. 2 Click the AI Agent node. Rewrite the system prompt with all four parts: ROLE (who am I?), CONSTRAINTS (what is off-limits?), STYLE (how do I sound?), OUTPUT FORMAT (how is the answer structured?). Keep it under 150 words. Example: 'You are LabBot, a specialist in protein folding. Only answer structural-biology questions — refuse clinical, legal, or political questions. Reply in 3-6 bullet points, no marketing language. End every answer with: ⚠ Uncertainty: .'
  3. 3 Pick ONE tool change: either (a) remove a tool you will not use, or (b) add a new HTTP Request Tool pointing at a public API you care about (PubMed, UniProt, a weather API — anything). Give it a clear description — the description is what the agent reads to decide when to call it.
  4. 4 Save (Ctrl+S) and open the chat.
  5. 5 Turn 1: ask a question inside your domain (e.g. 'What are the main steps of protein folding?'). Turn 2: ask a follow-up that only makes sense with memory (e.g. 'Which of those steps is most error-prone?'). The agent should answer turn 2 without you re-explaining what 'those steps' refers to — that proves the memory node is feeding previous turns back in.
  6. 6 Turn 3: ask something your constraints forbid (e.g. a clinical question). Confirm the agent refuses politely.
  7. 7 Take one screenshot of the new system prompt and one screenshot showing all three turns in the chat.
Deliverable

Your system prompt (showing the four parts) + screenshot of the two-turn memory test + screenshot of a refused out-of-scope question.

Self-check · tick before you mark done

✎What would you need to add to this workflow so the agent can only answer questions it can verify against a specific document? Keep that question in mind — we will answer it in session 03.