| name | learn |
| description | Build a focused 20-hour learning plan for a topic, structured as 10 two-hour sessions following the Pareto principle (20% that drives 80% of results). Use when the user wants to learn a topic fast, build a study plan, or create a structured curriculum. |
| argument-hint | <topic> |
Learn
Build a 20-hour plan to learn a topic fast, focused on the 20% that drives 80% of results. Output 10 two-hour sessions, each ending with a 15-minute review.
Step 1: Clarify the topic (if needed)
If the topic is ambiguous (e.g., "ML" → supervised? deep learning? deployment?), ask one clarifying question before proceeding. Otherwise skip.
Step 2: Research the 20%
Delegate to the researcher subagent to identify:
- The core concepts that experts agree drive most outcomes in this topic
- The best resources (books, courses, docs, videos, repos) — prefer canonical/authoritative sources
- Common beginner traps to skip
Use subagent_type: "researcher", model: "haiku".
Prompt:
text: What are the 20% of concepts in [TOPIC] that drive 80% of practical results? List the highest-signal resources (books, courses, official docs, key papers/repos) a motivated learner should use to master those concepts in ~20 hours. Flag common time-wasters to skip.
mode: deep-research
Step 3: Build the plan
Synthesize the research into 10 sessions. Each session:
- 2 hours total: ~105 min learning + ~15 min review
- One clear learning objective per session — a concrete capability, not a vague theme
- Progression: foundations → core mechanics → applied practice → integration
- Concrete resources: name the specific chapter, video, doc page, or exercise — not "read about X"
- Hands-on bias: at least half the sessions include building, coding, or solving problems
- Review block (last 15 min): recall without notes, summarize in own words, list 1-2 questions to revisit
Step 4: Present
Output format:
# 20-Hour Plan: [Topic]
**Focus:** [one-sentence framing of the 20% being targeted]
**Prerequisites:** [what learner should already know, or "none"]
---
## Session 1 — [Title] (2h)
**Objective:** [concrete capability after this session]
**Learn (105 min):**
- [Resource + specific section/chapter/exercise]
- [Resource + specific section/chapter/exercise]
**Review (15 min):**
- [Recall prompt or summary task specific to this session]
## Session 2 — [Title] (2h)
...
[continue through Session 10]
---
## Capstone check
After Session 10, you should be able to: [3-5 bullet outcomes]
## Sources
- [Title](url)
- [Title](url)
Constraints
- Do not pad sessions with filler. If the topic genuinely fits in fewer hours, say so and propose a shorter plan instead.
- Every resource must be named specifically (title + section). No "find a tutorial on X."
- The review block is non-negotiable — it's where retention happens.