quiz-me
Skill18 starsChanged 5 months agoMITView on GitHub
Quiz the user on code changes, a spec, or an implementation plan to verify genuine understanding before committing, building, or approving. Especially valuable when working with AI-generated code or spec-driven development. Use when the user says "quiz me", "/quiz-me", "test my understanding", "quiz me before I commit", "quiz me on this spec", "quiz me on the plan", "make sure I understand this", or anything suggesting they want to verify comprehension of a diff, design doc, or architectural plan.
18 other projects use a version of this.
Adopted by
Most recently taken up by
- jellydn/my-ai-toolstheir own version, first seen Nov 15, 2026
- yarikleto/claude-teacher-plugintheir own version, first seen Nov 14, 2026
- marnickvda/pop-quiztheir own version, first seen Nov 8, 2026
- whyashthakker/estacktheir own version, first seen Nov 8, 2026
- marnickvda/know-it-alltheir own version, first seen Nov 8, 2026
- matthlh/school-3-1their own version, first seen Nov 6, 2026
Used with
Projects with it also use reflect and handoff.
Other files people use alongside this one.
Files found in the same public projects as a version of this one, more often than they turn up anywhere else. Each project counts once, large collections count for less, and this file's own repository is left out.
Microsoft AzureSkillreflect
Learning capture system that extracts HIGH/MED/LOW confidence patterns from conversations to prevent repeating mistakes. Use after user corrections ("no", "wrong"), praise ("perfect", "exactly"), or when discovering edge cases. Complements .squad/agents/{agent}/history.md and .squad/decisions.md.
Also in 3 projects that use this
MicrosoftSkillhandoff
Compact the current conversation into a handoff document for another agent to pick up.
Also in 3 projects that use this
Other versions18 projects use a version of this.
Word for word · 0
None found.
Edited · 0
None found.
Wrote their own version · 18
- jellydn/my-ai-tools★ 129same name
- Ciucky/no-numb★ 32same name
- kevmoo/kevmoo_skills★ 23same name
- yarikleto/claude-teacher-plugin★ 15same name
- Ic3b3rg/learn-skills★ 9same name
- Rayhaan1345/socratic-tutor★ 5same name
- marnickvda/know-it-all★ 3same name
- marnickvda/pop-quiz★ 3same name
- bogdanbaciu21/skills★ 2same name
- matthlh/school-3-1★ 1same name
8 more not listed
How matches are found. Word for word: the exact same text. Edited: shares passages with this file (“99% same” means 99% of its five-word phrases also appear here; 10% or more counts). Their own version: similar meaning, or, weakest, the same name. We can’t tell which project had it first. Projects are counted among the 200 most-starred candidates, and mentions are links to the repository we found, up to 50 per site, so both counts are a floor.
The file
We split this skill into the rules it gives an agent
10 rules. 0 are common, 8 are distinctive, and 2 are very specific to this project.
0
Common
Most of the 18 edited versions we read keep these.
8
Distinctive
This author's own choices. Worth a look.
2
Very specific
Tied to their project's names, files or tools. Probably skip.
- Common
- Distinctive
- Very specific
- Distinctive: line 14
Code mode (default): Read git changes.
- Distinctive: line 16
Default: git diff HEAD (staged + unstaged)
- Very specific: line 24
Read the specified file path, or find the most recent file in docs/superpowers/specs/
- Very specific: line 29
Read the specified plan file, or look for recent plan files in docs/superpowers/plans/
- Distinctive: line 32
If ambiguous, ask the user which artifact to quiz on before proceeding.
- Distinctive: line 36
Read the artifact carefully.
- Distinctive: line 60
If [external dependency] isn't available, what breaks first?
- Distinctive: line 67
Wait for the answer
- Distinctive: line 76
On partial or wrong answers: ask one follow-up question that exposes the gap rather than explaining it - let them reason to the answer.
- Distinctive: line 84
If they still miss it after the follow-up, give a short correction and move on.
“Common” is measured: kept by at least 6 in 10 of the 18 edited versions of this file we could read. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.
Files
Quiz Me
Your job: read an artifact (code diff, spec, or plan) and quiz the user to verify genuine understanding - not explain it to them. This is active recall, not passive review. The goal is to surface gaps before they become bugs or wrong implementations.
Step 1: Determine source
Three modes, detected from context or user request:
Code mode (default): Read git changes.
- Default:
git diff HEAD(staged + unstaged) - "last commit" / "previous commit" →
git diff HEAD~1 HEAD - A specific SHA →
git diff <sha>~1 <sha> - "last N commits" →
git diff HEAD~N HEAD - A PR number → use gh MCP/plugin
Spec mode: Triggered when the user mentions a spec, design doc, or "before I implement".
- Read the specified file path, or find the most recent file in
docs/superpowers/specs/ - Questions target: data model decisions, interface contracts, scope boundaries, constraints
Plan mode: Triggered when the user mentions a plan, implementation plan, or "architectural decisions".
- Read the specified plan file, or look for recent plan files in
docs/superpowers/plans/ - Questions target: why this approach over alternatives, ordering of steps, dependencies between phases, risks
If ambiguous, ask the user which artifact to quiz on before proceeding.
Step 2: Generate 5–7 questions
Read the artifact carefully. Generate questions that require thinking, not just reading the artifact back. A good question cannot be answered by copy-pasting a line from the source.
For code diffs:
- Why was this approach chosen over a simpler/different one?
- What breaks if this line/block is removed?
- What happens with null, empty, or unexpected inputs?
- What does this migration change for existing data?
- What other parts of the system does this affect?
For specs:
- Why is [design decision X] made this way rather than [obvious alternative]?
- What's the contract of the provider protocol - what must every implementation guarantee?
- What's out of scope, and why?
- What happens if [constraint from spec] is violated?
- How does [component A] communicate with [component B]?
For plans:
- Why does Phase 1 end here rather than including [next thing]?
- Why was [architectural decision] chosen over [alternative you spotted]?
- What's the dependency between [step A] and [step B] - could they be reversed?
- What's the riskiest assumption in this plan?
- If [external dependency] isn't available, what breaks first?
Scale to what's in the artifact - a two-line fix warrants different depth than a full system design.
Step 3: Run the quiz
- Ask one question
- Wait for the answer
- Grade it - see Grading below
- Move to the next
- Final summary: X/Y + one observation
Grading
Correct = understands the why, not just the what. Partial credit when the right idea is there but an important nuance is missing. Don't accept vague answers - if they say "it's more efficient", ask them to be specific.
On partial or wrong answers: ask one follow-up question that exposes the gap rather than explaining it - let them reason to the answer. Only reveal the correction if they miss the follow-up too.
Good follow-up patterns:
- "You said X - what happens when [edge case Y]? Does your answer still hold?"
- "If that were true, what would [specific scenario] do?"
- "What does the spec say about [the thing they glossed over]?"
If they still miss it after the follow-up, give a short correction and move on. Not a lecture.
Tone
Tough but supportive senior dev. Not a gotcha machine - you want them to succeed. If they ace it, say so clearly.
Does it work?
Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.
No reviews yet
People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.
- It worked
- Partly
- It didn't
Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.
Your trail
Files you open will collect here, so you can find them again.
Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.
Now do this for your own.
Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.

