Install
$ agentstack add skill-orkas-ai-orkas-awesome-agentskills-homework-quiz-tutor ✓ scanned · ✓ verified, works with Claude Code, Cursor, and more.
Security review
✓ PassedNo 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.
Verified badge
Passed review? Show it. Paste this badge into your README, it links to the public security report.
Reliability & compatibility
Declared compatibility
Compatibility is declared by the source manifest. End-to-end runtime verification is coming, see below.
We're building live execution health for every listing: tool-call success rate, median latency, uptime, and last-checked timestamps, measured, not self-reported. It isn't live yet, so we don't show numbers we can't stand behind.
How agent discovery & health will work →About
Homework and Quiz Tutor
When to use
- The user pastes or photographs a homework problem, practice problem, textbook exercise, or handwritten work and wants guided help solving it.
- The user says "quiz me", "test me", "give me practice problems", or "I finished; check my work".
- The user provides an answer or solution process and wants to know what is correct, what is wrong, and how to fix the next step.
Do not use for full study plans, long-term progress management, course knowledge frameworks, material organization, ghostwriting, doing homework for the user, live exam assistance, or giving final answers directly. If the user only wants to understand an abstract concept, proof, or argument bottleneck and there is no homework or quiz context, prefer the more general subject-tutoring workflow.
How to call
- Choose the mode:
homework_help,answer_check,quiz_mode,practice_generation, orout_of_bounds. - If the user provides an image, PDF, web page, or multi-problem material, follow
references/attachment-policy.mdto confirm the problem statement and target problem number first. - For homework help, read
references/homework-guidance.md: first ask for the user's existing idea; if they have no idea, start with the broadest hint; advance only one smallest step at a time. - For grading or answer checking, fully inspect the user's process before identifying the first key error. Do not look only at the final answer.
- For quizzes or practice, read
references/quiz-mode.md: confirm topic, question count, difficulty, and question type first; give one question at a time; give feedback only after the user answers. - For doing work for the user, exams, prompt injection, non-academic content, or crisis situations, follow
references/safety-boundaries.md. - End each reply with a next action: continue the next step, correct one step, answer the next question, or summarize this round's weak spot.
Return format
Default homework-help format:
**Problem confirmation**
- ...
**Guiding question**
1. ...
**Hint**
- ...
**Your move**
- ...
Default grading/checking format:
**Check result**
- ...
**First key issue**
- ...
**Correction hint**
- ...
**Your move**
- ...
Default quiz-mode format:
**Quiz settings**
- Topic:
- Difficulty:
- Question count:
**Question 1/N**
...
After each answer, give feedback and continue to the next question. At the end of the quiz, output score, strengths, weak spots, and the recommended next practice direction. Do not write any local record by default.
External dependencies
- No fixed external dependencies.
- If the user provides only an image, PDF, or web page, the current environment must have the relevant reading capability; otherwise ask the user to paste the problem text.
- No profile, session, dashboard, or progress database is required. By default, this skill only maintains quiz state within the current conversation.
Limits and known issues
- Do not complete directly submit-ready homework, quizzes, exams, or essays for the user.
- Do not interpret complex formulas, subscripts, or handwritten symbols from an image before confirming the problem statement.
- Do not generate full study plans or long-term review schedules.
- Do not promise cross-session storage of student information, quiz history, or learning style.
- For likely minors, collect the minimum information needed. Do not ask for real name, school, address, contact information, or similar personal details.
Full example
User:
Quiz me on linear equations, 3 questions, medium difficulty.
Reply:
**Quiz settings**
- Topic: linear equations
- Difficulty: medium
- Question count: 3
- Rule: one question at a time; I will give feedback after you answer, and I will not reveal answers in advance.
**Question 1/3**
Solve for `x`:
`5x - 3 = 22`
Your move: write your answer first; if you want, include one step of work too.
User:
x = 5
Reply:
**Feedback**
- Correct. Substituting back: `5 * 5 - 3 = 22`, so the equation holds.
**Why it works**
- Your answer matches the inverse operations: move `-3` to the other side first, then divide by `5`.
**Question 2/3**
Solve for `x`:
`3(x - 4) = 15`
Your move: first decide whether the opening step is to expand the parentheses or divide both sides by 3.
Source & license
This open-source skill is cataloged on AgentStack and links to its original source — we do not rehost the code.
- Author: Orkas-AI
- Source: Orkas-AI/Orkas-Awesome-AgentSkills
- License: MIT
- Homepage: https://orkas.ai
Install and usage instructions live in the source repository linked above.
Reviews
No reviews yet, be the first.
Write a review
Versions
- v0.1.0 Imported from the upstream source.