AgentStack
SKILL verified MIT Self-run

Check Understanding

skill-yugash007-edu-agent-skills-check-understanding · by yugash007

Use when verifying whether the learner truly understands a concept through reasoning and application, not recall trivia.

No reviews yet
0 installs
14 views
0.0% view→install

Install

$ agentstack add skill-yugash007-edu-agent-skills-check-understanding

✓ scanned · ✓ verified — works with Claude Code, Cursor, and more.

Security review

✓ Passed

No issues found. Passed automated security review. · v0.1.0 How review works →

  • Prompt-injection patterns
  • Secret / credential exfiltration
  • Dangerous shell & filesystem operations
  • Untrusted network calls
  • Known-malicious package signatures

What it can access

  • Network access No
  • Filesystem access No
  • Shell / process execution No
  • Environment & secrets No
  • Dynamic code execution No

From automated source analysis of v0.1.0. “Used” means the capability is present in the source — more access means more to trust, not that it’s unsafe.

Are you the author of Check Understanding? Claim this listing to set pricing, connect Stripe payouts, and keep 70% of every sale.
Sign up to claim

About

Purpose

Assess conceptual and practical understanding using reasoning-first prompts, then adapt teaching based on detected weak areas.

Activation

  • Concept explanation just completed. Learner claims understanding and needs validation. Repeated related mistakes. Readiness check needed before advancing.
  • Skip if: no concept context established, or user explicitly declines assessment.
  • Routing: pair with teach-concept for explanation→assessment loops. Escalate to socratic-mode when misconceptions persist.

Inputs

  • Target concept(s), learner level, prior mistakes/confusion signals, repo/project context.

Workflow

  1. Target — Select 1–2 concepts being validated.
  2. Question — Generate tiered questions: conceptual reasoning, practical application, debugging/diagnostic.
  3. Evaluate — Grade reasoning quality, not keyword match.
  4. Classify — Categorize mistakes: misconception, partial model, or execution gap.
  5. Correct — Explain root cause and provide corrected model.
  6. Recheck — One follow-up question to confirm recovery.

Rules

  • DO: test reasoning and transfer, not memorization.
  • DO: use plausible distractors in MCQ format.
  • DO: explain why an answer fails, not just that it's wrong.
  • DO: track weak areas across turns when context allows.
  • DON'T: use only right/wrong labels — always output mistake type and remediation.
  • DON'T: end without a recheck after correction.
  • DON'T: ignore repeated weak areas — log and prioritize them.

Output

Responses should contain: concepts under assessment, questions (conceptual + practical + diagnostic), evaluation (strengths, weak areas, mistake types), corrective feedback, and recheck question. Format naturally.

Checklist

  • [ ] Includes conceptual and practical checks.
  • [ ] Mistakes categorized, not just scored.
  • [ ] Corrective feedback explains root cause.
  • [ ] Follow-up recheck present.

Source & license

This open-source skill is cataloged on AgentStack and links to its original source — we do not rehost the code.

Install and usage instructions live in the source repository linked above.

Reviews

No reviews yet — be the first.

Versions

  • v0.1.0 Imported from the upstream source.