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 skills/lucasheriques/shipmate/teaching-designnpx skills add lucasheriques/shipmate --skill teaching-designgit clone --depth 1 https://github.com/lucasheriques/shipmateWrote this? Show the measurements
A badge with what this costs and how it scanned, read live from this page, so it follows the numbers instead of freezing them. Markdown for a README, HTML for a documentation site or a project page.
[](https://agentmods.dev/skills/lucasheriques/shipmate/teaching-design)<a href="https://agentmods.dev/skills/lucasheriques/shipmate/teaching-design"><img src="https://agentmods.dev/badge/skills/lucasheriques/shipmate/teaching-design.svg" alt="Measured on agentmods" height="20"></a>What 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 5.1 | $0.00066 | $0.01859 |
| Opus 5 | $0.00033 | $0.00929 |
| Sonnet 5 | $0.00013 | $0.00372 |
| Haiku 4.5 | $0.00007 | $0.00186 |
Grade A, and why
teaching-design 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 6d 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 — 77 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Teaching Design
When to use
Use this whenever the deliverable teaches: a workshop, course, talk, video script, webinar, stream, or onboarding session. Also use it as a diagnostic when existing content feels flat — every failure to learn traces back to a skipped step below. Core stance: energy, attention, and learning are properties of the design, not the delivery. Never open slides or start scripting before finishing phases 1-3.
The procedure
Phase 1 — Learning outcomes (what)
- Profile the audience first: who are they, how experienced, why are they showing up, what are they skeptical about? This decides what to cut. A session about "everything" is always wishy-washy.
- Write Learning Outcomes: the specific delta between what the audience believes/can do on arrival vs departure. Sharpen each with: "What's the specific argument I'm making? What do they believe differently at the end?"
- Reject topics; demand takeaways. "CVs" is a topic. "What to delete from your CV to make it stronger" is an outcome. First passes are always too vague — rewrite.
- Cut ruthlessly. You are not a Wikipedia article; the designer is a curator, and deciding what's out matters as much as what's in.
- Expand each outcome into 2-4 supporting ideas: "What else must they know, believe, or be able to do to absorb the main outcome?"
- Get feedback on the outline now — before anything is built, while changes are free. The outline doubles as marketing copy.
Phase 2 — Skeleton (when)
- Place breaks BEFORE designing content: 15 min coffee / 60 min lunch after every 60-90 minutes. Breaks are sacred — never shrink one to cram content.
- Budget 30-45 minutes per Learning Outcome (a 30-min outcome = ~10 min lecture + 10 min exercise + 5 min follow-up + 5 min Q&A). So: 2-3 outcomes per 90-minute chunk, ~2 per short video, ~3 per talk.
- Jam the outcomes into the schedule and sanity-check. If it's crammed, cut a whole outcome NOW — an ignored squeeze resurfaces on the day as rushing, skipped exercises, and drained energy.
- Secretly plan to end early. Nobody in history has been upset about ending 15 minutes early.
- The Skeleton (audience + schedule + outcomes) fits on one page. Keep it at hand while delivering.
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.
- 6d ago First seen · 77 lines · 66 tokens per session scan A d713492f853a
teaching-design is a skill published in the GitHub repository lucasheriques/shipmate (4 stars, last pushed 1mo ago), licensed MIT. It adds 66 tokens to every session and 1,859 once invoked, about $0.0003 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 skills, from other repositories
solo-failures
Understand why startups fail — the most common failure modes, anti-patterns, and how to avoid them, backed by data from 101 founder interviews.
solo-roast
Get your startup plan roasted with data — Garry Tan style.
solo-analyze
Evaluate a startup idea against real patterns from 101 founder interviews.
solo-growth
Get data-backed growth and marketing recommendations for your business.
solo-patterns
Browse and search the startup pattern database (11 categories, 80+ patterns).
solo-playbook
Get a data-backed step-by-step playbook for any startup goal.