| name | docker-python-lab |
| description | Generates an interactive Docker Python lab block for MkDocs textbook pages, where students write and run real Python code inside an isolated Docker container. Use this skill whenever someone asks to add a Python lab, code runner, interactive Python exercise, or runnable code block to a textbook page that uses Docker (not Skulpt). Also use it when adding multiple labs to a single page, setting up the shared CSS/JS infrastructure, or creating a timing/benchmark lab that shows students how long each phase takes. Always invoke this skill instead of writing docker lab HTML by hand. |
Docker Python Lab Generator
What This Skill Does
Generates interactive Python lab blocks for MkDocs pages in the learning-python
textbook (or any MkDocs project following the same pattern). Each lab shows a
code editor, Run and Reset buttons, and an output area. Code runs inside a fresh,
isolated Docker container via a local HTTP service on port 5001.
This is the Docker counterpart to the Skulpt lab pattern — same look and feel,
but students run real Python (not browser-emulated Python) with access to the full
standard library and third-party packages.
Prerequisites: Shared Infrastructure
Before generating any lab block, make sure the shared files exist. Check once
per session; skip if already in place.
1. Check / create docs/css/docker-lab.css
If the file is missing, copy from the skill's assets:
cp ~/.claude/skills/docker-python-lab/assets/docker-lab.css \
docs/css/docker-lab.css
2. Check / create docs/js/docker-lab.js
cp ~/.claude/skills/docker-python-lab/assets/docker-lab.js \
docs/js/docker-lab.js
3. Add to mkdocs.yml (if not already present)
Check extra_css and extra_javascript in mkdocs.yml. If the entries are
missing, add them:
extra_css:
- css/docker-lab.css
extra_javascript:
- js/docker-lab.js
Do not duplicate entries that are already there.
4. Verify the service script exists
The runtime service lives at scripts/run-python-docker.sh. If it is missing
from the project, tell the user — the labs won't work without it.
Lab HTML Template
Each lab needs a unique suffix — a short string that distinguishes it from
other labs on the same page. Use "1" for the first lab, "2" for the second,
and so on. The suffix appears in every element ID so the JS can find the right
elements.
Standard lab (text output only)
<div id="docker-lab-SUFFIX">
<div id="docker-editor-SUFFIX">
<textarea id="docker-code-SUFFIX" spellcheck="false">PYTHON_CODE_HERE
</textarea>
<div id="docker-buttons-SUFFIX">
<button id="docker-run-SUFFIX" onclick="runDocker('SUFFIX')">▶ Run</button>
<button id="docker-reset-SUFFIX" onclick="resetDocker('SUFFIX')">↻ Reset</button>
</div>
<pre id="docker-output-SUFFIX" class="docker-output">Output will appear here after you click Run.</pre>
</div>
</div>
Rules:
- Replace every
SUFFIX with the same string (e.g., 1, 2, hello-world).
- Put the Python code directly inside the
<textarea> tag — no extra indentation
relative to column 0, because leading spaces become part of the code.
- The
</textarea> closing tag must be on its own line with no trailing spaces
before it; any blank line between the code and </textarea> adds an empty line
in the editor.
- Do not add
rows="N" — the JS auto-sizes the textarea to fit the code.
- Do not add
id="main" to any element — that conflicts with the p5.js canvas
parent convention used elsewhere in this project.
Timed lab (shows phase-by-phase timing breakdown)
Only add this variant when the lesson is specifically about how Docker execution
works. It requires suffix "4" by convention and adds a timing table below the
buttons.
<div id="docker-lab-4">
<div id="docker-editor-4">
<textarea id="docker-code-4" spellcheck="false">PYTHON_CODE_HERE
</textarea>
<div id="docker-buttons-4">
<button id="docker-run-4" onclick="runDockerTimed()">▶ Run + Time</button>
<button id="docker-reset-4" onclick="resetDockerTimed()">↻ Reset</button>
</div>
</div>
<div id="docker-timing-display" style="display:none; margin-top:12px;">
<table id="docker-timing-table">
<thead>
<tr><th>#</th><th>Phase</th><th style="text-align:right">Time (ms)</th><th>Bar</th></tr>
</thead>
<tbody>
<tr><td>1</td><td>Send to service (network)</td>
<td id="td-network-send" style="text-align:right">—</td>
<td><div class="timing-bar" id="bar-network-send"></div></td></tr>
<tr><td>2</td><td>Container startup</td>
<td id="td-startup" style="text-align:right">—</td>
<td><div class="timing-bar" id="bar-startup"></div></td></tr>
<tr><td>3</td><td>Python execution</td>
<td id="td-exec" style="text-align:right">—</td>
<td><div class="timing-bar" id="bar-exec"></div></td></tr>
<tr><td>4</td><td>Return to browser (network)</td>
<td id="td-network-return" style="text-align:right">—</td>
<td><div class="timing-bar" id="bar-network-return"></div></td></tr>
<tr style="font-weight:bold; border-top: 2px solid #642580;">
<td colspan="2">Total round-trip</td>
<td id="td-total" style="text-align:right">—</td>
<td></td></tr>
</tbody>
</table>
<p id="docker-timing-note" style="font-size:0.85em; color:#666; margin-top:6px;"></p>
</div>
<pre id="docker-output-4" class="docker-output" style="margin-top:10px;">Output will appear here after you click Run + Time.</pre>
</div>
Complete Page Pattern
A page with three labs looks like this:
# Page Title
Intro text explaining what the page covers and that students must run
`bash scripts/run-python-docker.sh` in a separate terminal first.
## Lab 1 — Hello, World!
Brief explanation of what this lab demonstrates.
<div id="docker-lab-1">
<div id="docker-editor-1">
<textarea id="docker-code-1" spellcheck="false">print("Hello, World!")
</textarea>
<div id="docker-buttons-1">
<button id="docker-run-1" onclick="runDocker('1')">▶ Run</button>
<button id="docker-reset-1" onclick="resetDocker('1')">↻ Reset</button>
</div>
<pre id="docker-output-1" class="docker-output">Output will appear here after you click Run.</pre>
</div>
</div>
**Try these experiments:**
- Change the message and run again.
---
## Lab 2 — Variables
<div id="docker-lab-2">
<div id="docker-editor-2">
<textarea id="docker-code-2" spellcheck="false">x = 42
print("The answer is", x)
</textarea>
<div id="docker-buttons-2">
<button id="docker-run-2" onclick="runDocker('2')">▶ Run</button>
<button id="docker-reset-2" onclick="resetDocker('2')">↻ Reset</button>
</div>
<pre id="docker-output-2" class="docker-output">Output will appear here after you click Run.</pre>
</div>
</div>
Choosing Good Lab Code
- Keep starter code short (5–15 lines) so students can read it at a glance.
- Every lab should produce visible output — at least one
print() call.
- Avoid
input() — there is no stdin in the Docker runner.
- Avoid file I/O — the container has no persistent filesystem.
- Avoid
import of non-stdlib packages unless python:3.11-alpine includes
them (it ships only the stdlib).
- Match the "See It — Run It — Modify It" rhythm from the Skulpt labs:
starter code runs and shows something, then "Try these experiments" gives
specific, achievable modifications.
Service Not Running — User-Facing Error
If a student clicks Run before starting the service, the lab shows:
Cannot connect to the Python Docker service.
Please open a terminal and run:
bash scripts/run-python-docker.sh
Then reload this page and try again.
No extra error handling is needed in the page markdown — docker-lab.js handles
this automatically.
Multiple Labs per Page — Checklist
When generating labs for a single page: