awesome-exam-skills

A curated collection of Claude skills for exam preparation and academic intelligence.

---

Each skill in this repository is a structured system prompt engineered to give Claude repeatable, well-defined exam-prep capabilities. Drop any skill into Claude via Projects, the API, or any compatible skill runner — no setup beyond copy-paste.

Skills

Skill	Description	Exam Types	Status
pyq-analyzer	Analyzes previous year question papers to extract patterns, score confidence, flag anomalies, and produce a prioritized study plan	T1–T5 (all subjects)

See ROADMAP.md for planned additions.

---

Getting Started

Every skill follows the same installation pattern.

Claude.ai Projects (recommended)

1. Navigate to claude.ai and open or create a Project
2. Under Project Instructions, paste the full contents of the skill's SKILL.md
3. Follow the skill-specific usage instructions in its README.md

API

import anthropic

with open("skills/pyq-analyzer/SKILL.md", "r") as f:
    skill = f.read()

client = anthropic.Anthropic()
response = client.messages.create(
    model="claude-sonnet-4-20250514",
    max_tokens=4096,
    system=skill,
    messages=[{"role": "user", "content": "I've uploaded 5 DBMS papers. Analyze them."}]
)

Other Environments

Any tool that accepts a Claude system prompt works — agent frameworks, custom UIs, local runners. Drop the SKILL.md content in as the system prompt.

---

Repository Structure

awesome-exam-skills/
├── README.md               ← You are here
├── CONTRIBUTING.md         ← Contribution guidelines
├── ROADMAP.md              ← Planned skills and proposals
├── CODE_OF_CONDUCT.md      ← Community standards
├── LICENSE
└── skills/
    └── pyq-analyzer/
        ├── SKILL.md        ← The skill (paste this into Claude)
        ├── README.md       ← Usage, examples, limitations
        └── examples/
            └── README.md   ← Sample outputs

New skills each get their own folder under skills/. See CONTRIBUTING.md for the required structure.

---

Design Philosophy

All skills in this repository are built around a single principle:

Show raw data. Let the student decide.

No skill will tell a student to ignore a topic or assert that something will or won't appear. Skills surface historical patterns, quantify confidence, and flag anomalies — then get out of the way. The student makes the call.

---

Contributing

Contributions are welcome — whether you're reporting an edge case, improving an existing skill, or proposing an entirely new one.

See CONTRIBUTING.md for full guidelines. The short version:

• Bug reports / edge cases → open a GitHub issue
• Improvements to an existing skill → fork, edit SKILL.md, test on real papers, open a PR
• New skill proposals → open an issue with the skill-proposal label before building

---

License

MIT — use it, fork it, modify it, build on it.

---

Built by Abdul Fattah · LinkedIn

If this project helped you, consider starring the repository.