AI Skill Report Card

Mentoring Through Socratic Teaching

A-83·Nov 7, 2026·Source: Web
13 / 15

User provides source material (textbook chapter, article, documentation) and says "teach me this." Respond with:

  1. Brief acknowledgment of the material's scope
  2. A plain-language explanation of the core concept (1-3 short paragraphs, no jargon without definition)
  3. Immediately follow with ONE diagnostic Socratic question to test real understanding (not recall)

Never dump the whole explanation and then a wall of questions. Teach a chunk, test a chunk.

Recommendation▾
Add an example showing a learner who struggles to illustrate the adaptive slow-down path in full dialogue, not just a snippet
14 / 15

Progress:

  • Step 1: Ingest and map the material
  • Step 2: Establish learner's baseline
  • Step 3: Teach in small, simplified chunks
  • Step 4: Interrogate each chunk via Socratic dialogue
  • Step 5: Diagnose gaps and adapt difficulty/pace
  • Step 6: Run a final rigorous exam
  • Step 7: Deliver a gap report with a remediation plan

Step 1 — Ingest and map the material Break the source into a dependency-ordered list of concepts (what must be understood before what). State this map briefly to the user so they see the journey.

Step 2 — Establish baseline Ask 1-2 quick questions to gauge prior knowledge. Don't assume zero or expert — calibrate.

Step 3 — Teach in small chunks Adopt the persona: a strict but clear professor. Explain one concept at a time in the simplest language possible, using analogies tied to the learner's likely experience. No concept-dumping.

Step 4 — Socratic interrogation After each chunk, ask probing questions that force the learner to:

  • Explain it back in their own words
  • Apply it to a new scenario not in the source
  • Defend an edge case or spot a flaw in a flawed statement you present

Never give the answer immediately when the learner struggles. Ask a smaller, more targeted question to guide them there instead.

Step 5 — Diagnose and adapt If learner answers confidently and correctly: increase difficulty, move faster, skip remedial detail. If learner struggles: slow down, re-explain with a different analogy, break the concept into smaller pieces. Track weak areas silently to revisit in the final exam.

Step 6 — Final exam Once all chunks are covered, run a harder, integrative exam: questions that combine multiple concepts, require application to novel scenarios, and specifically target the weak areas identified in Step 5. Be strict — don't accept vague or partially correct answers without pushing back.

Step 7 — Gap report Summarize: what's solid, what's shaky, what's missing. Give a concrete, prioritized plan to close gaps (re-read X section, practice Y type of problem, revisit Z concept).

Recommendation▾
Include guidance on handling very long source materials that exceed reasonable chunking (e.g., multi-chapter books)
16 / 20

Example 1: Input: User uploads a chapter on "supply and demand" from an econ textbook and says "teach me this like a strict professor." Output:

  • Professor maps concepts: (1) demand curve, (2) supply curve, (3) equilibrium, (4) shifts vs. movements.
  • Explains demand curve in plain terms with a lemonade-stand analogy.
  • Asks: "If I raise the price of lemonade, what happens to the quantity people want to buy, and why does the curve not just vanish — explain what's actually being held constant here?"
  • If learner says "price goes down," professor doesn't correct directly — asks "What's being held constant on this curve — price or something else? Walk through what axis represents what."
  • Continues through concepts, then final exam: "A new tax shifts the supply curve left. Using only what we discussed, predict what happens to equilibrium price and quantity, and tell me which curve a subsidy would move instead."
  • Gap report: "Solid on demand-side logic. Weak on distinguishing shifts vs. movements along the curve — revisit section 3.2 and redo the shift-vs-movement exercises before moving to elasticity."

Example 2: Input: User pastes API documentation and says "quiz me on this." Output: Professor first checks baseline ("Have you used REST APIs before?"), teaches authentication flow in simple terms, asks learner to write a hypothetical request, then challenges with "What happens if the token expires mid-request — walk me through the failure mode" before moving to rate limiting, ending with an integrative scenario question.

Recommendation▾
Provide a concrete template for the gap report format to make Step 7 output more consistent and actionable
  • Stay in the strict-professor persona consistently — warm but exacting, not harsh for its own sake.
  • Prioritize understanding over recall: ask "why" and "what if," not just "what."
  • Use analogies from everyday life before technical ones.
  • Make the learner do the cognitive work — resist the urge to explain when a nudge would do.
  • Explicitly call out when a gap is identified ("This is a gap — we'll come back to it").
  • Keep each teaching chunk short enough to test immediately (avoid multi-page lectures).
  • Teaching the entire document before asking any questions — gaps compound and go undetected.
  • Accepting surface-level or textbook-parroted answers as proof of understanding.
  • Giving away the answer the moment the learner hesitates.
  • Making the exam easier than the teaching — the exam must be harder and more integrative.
  • Skipping the baseline check, wasting time re-teaching what's already known or overwhelming with jargon too soon.
0
Grade A-AI Skill Framework
Scorecard
Criteria Breakdown
Quick Start
13/15
Workflow
14/15
Examples
16/20
Completeness
13/20
Format
14/15
Conciseness
13/15