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/xiaolai/mac-it-guy-pro/learngit clone --depth 1 https://github.com/xiaolai/mac-it-guy-proWhat 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 | $0.00024 | $0.01188 |
| Opus 5 | $0.00012 | $0.00594 |
| Sonnet 5 | $0.00005 | $0.00238 |
| Haiku 4.5 | $0.00002 | $0.00119 |
Grade A, and why
learn 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 — 65 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Learn — the model, not the manual
Read ${CLAUDE_PLUGIN_ROOT}/skills/tutoring/SKILL.md first — the two modes, the grounding rule, the ten-section structure, the dating rule, and the language rule are binding. Read ~/ITGuy/machine.md if it exists, for the user's language preference and for real numbers to teach from.
The user does not need to memorise commands. Build the model that lets them decide — when to worry, what to check first, what to buy, what to ignore, what to automate next.
Resolve from $ARGUMENTS: why · a topic · list · empty.
empty — offer what their machine has already taught them
Do not ask "what would you like to learn?" A user who could answer that does not need this command. Instead, derive candidates from what they have actually met:
- Read
~/ITGuy/visits.log(what was done),ledger.jsonl(what was believed and revised), the profile's Live Conclusions and Watch List, andtoolbox.json(what they automated). - Turn each into a candidate phrased as their experience, not a subject name: "Why your Downloads folder keeps filling up" rather than "File system management." "What that Wi-Fi measurement actually meant" rather than "802.11 fundamentals."
- Offer at most four with AskUserQuestion, plus an "something else" option that accepts a free-text topic.
- Mark anything already in
~/ITGuy/learn/as a re-read rather than a new map.
If there is no history at all, say so honestly and offer the three domains this plugin covers — keeping your Mac healthy, your home network, and automating chores — as a starting menu.
why — just-in-time, the mode that matters most
Explain what just happened and why it was the fix. Scope is the current session and the last entries in visits.log.
- One or two paragraphs. No sections, no diagrams, no reading assignment.
- Ground it in the actual numbers that were measured, not in an abstraction.
- Say what the user should now recognise next time — the signal, not the procedure.
- Close with one offer at most: a full map on the topic, if one would genuinely add something.
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 · 65 lines · 24 tokens per session scan A 03db63a4ce72
learn is a command published in the GitHub repository xiaolai/mac-it-guy-pro (5 stars, last pushed 1mo ago), licensed MIT. It adds 24 tokens to every session and 1,188 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
dashboard
Launch the learning dashboard web UI to view and edit plans, progress, and spaced repetition data.
journal
Write a journal entry for today's work session on the rem project.
evolve
Cluster hardened instincts into a proposed Command / Skill / Agent (Instinct Engine).
discovery
Step 0 — converges a vague intuition into a clear problem (5-whys + JTBD + why-now), teaching each framework as it goes, and seeds clues for persona/journey/metrics. The pre-PRD discovery step.
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.