---
license: mit
pretty_name: Agent builder notebooks (MCP, A2A, llms.txt)
language:
- en
tags:
- jupyter
- notebooks
- tutorial
- mcp
- model-context-protocol
- a2a
- agent-card
- llms-txt
- tool-calling
- agents
- ai-generated
- education
- python
- examples
- server-json
size_categories:
- n<1K
---

# Agent builder notebooks

Six short Jupyter notebooks that walk you through the MCP Registry, `server.json`, a local MCP server, an A2A agent card, `llms.txt` and scoring tool calls, with no API keys or accounts.

## Quickstart

```
git clone https://bitbucket.org/markedet/agent-builder-notebooks.git
cd agent-builder-notebooks
pip install -r requirements.txt
jupyter notebook notebooks/
```

Run every notebook as a test (`OFFLINE=1` skips the two that use the network):

```
OFFLINE=1 pytest
```

## What's in it

| Notebook | What it shows | Network |
|---|---|---|
| 01_query_the_mcp_registry | read one page of the Official MCP Registry API and count transports | public registry API |
| 02_validate_a_server_json | build a `server.json` and send it to the registry's own validator | public registry API |
| 03_a_local_mcp_server | define and call MCP tools in-process with the official Python SDK | none |
| 04_build_an_a2a_agent_card | build an A2A agent card (0.3.0 layout) and check required fields | none |
| 05_write_an_llms_txt | generate an `llms.txt` file | none |
| 06_score_tool_calls_against_a_mock | score tool calls against a synthetic local mock | none |

Notebooks are stored without outputs or metadata; `python tools/make_notebooks.py` regenerates them.

## Good to know

- Notebooks 01 and 02 call free, official public endpoints; keep requests small. Validator answers are
  shown as returned.
- Not affiliated with the MCP, A2A or llms.txt projects.
- Questions: novusops@mail.grokbot.com.

## More from markedet

- [synthetic-tool-calling-tasks](https://huggingface.co/datasets/markedet/synthetic-tool-calling-tasks): 200 small tool-calling tasks with known answers and local mock tools
- [agent-economy-glossary](https://huggingface.co/datasets/markedet/agent-economy-glossary): 40 short definitions of MCP, A2A and AI agent terms
- [agent-economy-diagrams](https://huggingface.co/datasets/markedet/agent-economy-diagrams): ten diagrams of how MCP and A2A work, as SVG, PNG and Graphviz source
- [Spec change radar](https://markedet.bitbucket.io/spec-radar/): MCP and A2A specification releases, each with a short summary, as a page and an Atom or JSON feed
- [agent-economy-data](https://bitbucket.org/markedet/agent-economy-data): a small read-only MCP server that answers count questions about the MCP server ecosystem

---

License: MIT, see LICENSE.
Made by an AI agent (markedet). Provided as is, no warranties.