Getting it into your agent
One page per mod, every tool's command on it. A separate URL per tool would split the same page into five that compete with each other.
npx agentmods add commands/mktoronto/python-clean-architecture/learn-architecturegit clone --depth 1 https://github.com/MKToronto/python-clean-architectureWhat it costs to keep this loaded
Counted locally with the o200k_base tokenizer, which is exact for GPT models; Claude uses its own tokenizer and its counts differ. Treat this as one consistent yardstick across the catalogue rather than a bill. Prices are per million input tokens.
| Model | Per session | Once invoked |
|---|---|---|
| Fable 5 | $0.00025 | $0.02407 |
| Opus 5 | $0.00013 | $0.01203 |
| Sonnet 5 | $0.00005 | $0.00481 |
| Haiku 4.5 | $0.00003 | $0.00241 |
Grade A, and why
learn-architecture scanned grade A with 0 findings against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured 2d ago.
A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.
Nothing flagged
None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.
How it starts
The opening of the file, as written. The whole thing — 123 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Teach the user an architecture or design topic from this skill's knowledge base. This command explains and deepens understanding — it does not refactor, review, or scaffold. The goal is for the user to understand both the theory and how it works in practice, including the trade-offs and comparisons an interviewer (or a thoughtful teammate) would probe.
The user may have given a topic in $ARGUMENTS (e.g. strategy, dependency injection, the 7 principles, drill me on repository). It may also contain a mode hint like drill or lesson.
Hard rule — this command never edits
This is a read-only, teaching command. It reads the user's code and the bundled example to explain — it must never create, modify, or delete any file as part of teaching. Showing "before/after" snippets, pointing at a violation, or saying "here's how you'd fix it" is explanation, not editing — present it in your reply, never apply it.
The only exception: the user explicitly asks you to apply a change to their code during the session (e.g. "go ahead and refactor that file", "apply that fix"). Even then:
- Stop and confirm first. Use AskUserQuestion to show exactly which file(s) and what change, and get an explicit yes before any edit. A vague "sure" or momentum from the lesson is not consent — confirm the specific change.
- This confirmation is mandatory even in auto-accept / auto-edit mode. Do not rely on the permission prompt to gate the edit — proactively ask via AskUserQuestion yourself, because that prompt may be suppressed. No silent edits, ever.
- Prefer to redirect. Applying refactors is the job of the action commands. Point the user to
/make-pythonic,/extract-god-class,/decouple,/add-endpoint, or/scaffold-apiinstead — only edit from this command if they decline the redirect and still want it here.
Step 1 — Pick the mode
Unless the arguments already make it obvious, use AskUserQuestion to ask "How do you want to learn this?" with two options:
What this file has done since we first saw it
Hashed on every crawl. A supply-chain change to an agent config is a question of when, not whether, so the history is kept rather than the latest state alone.
- 2d ago First seen · 123 lines · 25 tokens per session scan A 13a3efd8c450
learn-architecture is a command published in the GitHub repository MKToronto/python-clean-architecture (8 stars, last pushed 2mo ago), licensed MIT. It adds 25 tokens to every session and 2,407 once invoked, about $0.0001 per session on Opus 5. A static security scan graded it A with 0 findings. No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-31.
Other commands, from other repositories
getting-started
Interactive 10-minute guided tour for new Claude Craft users.
dashboard
Launch the learning dashboard web UI to view and edit plans, progress, and spaced repetition data.
learn
Start learning a new topic — asks clarifying questions, researches resources, and creates a structured learning plan.
quiz
Quiz yourself on a topic from your learning plan with adaptive difficulty and mixed question formats.
resources
Find curated learning resources — books, courses, tutorials, and docs for any topic.
review
View your learning progress — quiz scores, weak areas, and what to study next.