Learn Anything — Tutor-Style Learning Workflow (导师式带学)
Tutor-style learning workflow: diagnose goal → break into skills → teach with minimal examples → project task → mastery check → durable study files. 5-level mastery rubric, 6 prompt templates, 6 fill-in templates.
Free
Works with the AI Agents you already use
Learn Anything — Tutor-Style Learning Workflow (导师式带学)
Example session with this skill installed
帮我学线性代数前三章:先给我四周学习计划,每次学完做掌握度检查,错题记进错题本。
- Read your context and instructions
- Compiled the learn anything
Creates study/linear-algebra-learning-plan.md (4-week plan with weekly deliverables and acceptance criteria). Per session: teaches one skill point with a minimal example, quizzes one question at a time, then updates study/linear-algebra-mastery-check.md (level + evidence + weak spots) and study/linear-algebra-mistakes-log.md.
Connects securely to your tools. The creator never sees your data.
About this skill
What this is
A tutor-style learning workflow: the agent becomes a Chinese-speaking tutor + project coach. It diagnoses your goal and starting point, breaks the topic into 2-5 verifiable skills, teaches with minimal examples, assigns a small project-style task, checks mastery by quizzing (recall / fix-the-error / variant), and writes everything into a durable study/ archive instead of letting it scroll away in chat.
What's inside
- 9-step teaching flow: goal → starting point → skill breakdown → core principles → minimal example → project task → mastery check → file output → next step
- 5-level mastery rubric: 知道 / 会用 / 会改 / 会设计 / 会迁移 (know → use → modify → design → transfer), judged from at least two types of evidence, never from "I think I got it"
- Source strategy: official docs and primary sources first; facts vs. teaching judgment clearly separated; states which authorities it fell back on when no official source exists
- 6 prompt templates you can paste into any AI: learning ladder, top-20% plan, examiner one-question-at-a-time, one-page cheatsheet, resource filter, Feynman recall
- 6 fill-in templates: learning plan, study notes, session review, project brief, mastery check, mistakes log
- Output contract tuned for learning: conclusion-first, bounded steps, progress restated each round, no filler, concrete time estimates
Honest positioning
Validated by daily use across a full university semester (linear algebra, machine-learning math, discrete math, classics reading, and more) — the teaching flow, mastery checks and study-file conventions were refined in real tutoring sessions. It is built for a human who wants to actually learn: it makes you produce, recall and get quizzed instead of only reading explanations.
Chinese-first: explanations default to Chinese with English terms kept; ask for English to switch. Non-coding subjects (math, history, languages, writing) get output-style projects — short reports, explainer scripts, knowledge maps — no forced coding.
Teaches in Chinese by default (ask for English to switch). Mastery judgements rely on your answers to recall/variant questions — it quizzes you rather than reading your mind. File outputs need a writable working directory; without one it teaches chat-only and tells you what it would have written.
How to use
- 1
Connect your agent to the Agensi MCP
Copy the prompt and paste it into Claude, ChatGPT, Cursor or any agent that supports MCP. It sets up the connection for you. Manual setup
- 2
Sign in to Agensi
Your agent shows a sign-in link. Sign in and click Allow.
- 3
Ask for what you need
Your agent loads this skill and uses it on your task. It always asks before any payment.
Reviews
No reviews yet
Be one of the first to try it. Every listed skill goes through the trust checks below.
Security scanned
Checked by our 8-point scan before listing
Fresh listing
Recently published to Agensi
Free forever
No account required to browse
Trust & safety
Security scan
Security scan passed 9 days ago
- Free with an Agensi account
Permissions
No special permissions declared or detected