A collection of Claude Agentic Skills for intelligent document processing, study material generation, and PDF extraction.

This repository contains specialized skills that enable Claude to process documents and generate study materials. Skills include PDF-to-markdown conversion, Anki flashcard generation, and revision notes creation.

claude-skills/
├── README.md
├── extracting-pdfs/                    # PDF extraction skill (Active)
│   ├── SKILL.md                        # Skill definition & workflow
│   ├── cleanup-patterns.md             # Reference: noise patterns to remove
│   ├── image-handling.md               # Reference: processing extracted images
│   ├── sentence-reflow.md              # Reference: fixing fragmented text
│   └── table-formatting.md             # Reference: reconstructing malformed tables
├── anki-flashcard-generator/           # Anki flashcard generation skill
│   └── SKILL.md
├── revision-notes-generator/           # Revision notes generation skill
│   └── SKILL.md
└── archive/                            # Archived/legacy skills
    ├── pdf-extract/                    # PDF extraction (development version)
    │   ├── SKILL.md
    │   ├── extract_pdf.py              # Core Python extraction script
    │   ├── cleanup-patterns.md
    │   ├── image-handling.md
    │   ├── sentence-reflow.md
    │   └── table-formatting.md
    └── pdf-to-markdown-converter/      # Legacy skill (Deprecated)
        └── SKILL.md

The primary skill for extracting PDF content to clean, organized markdown format. This is the production-ready version with a sophisticated multi-step workflow.

Trigger: When a user uploads a PDF and wants to convert it to markdown.

Workflow:

1. Extract — Run Python script to get raw content and metadata
2. Analyse — Review extracted content for patterns and issues
3. Clean — Manually rewrite to remove noise (no automated scripts)
4. Organise — Apply formatting with proper heading hierarchy
5. Output — Deliver clean markdown with images

Key Features:

• Dual extraction methods with automatic fallback
  • pymupdf4llm: Primary method for better markdown/table formatting
  • pymupdf: Fallback for scanned/image-based PDFs
• Comprehensive image extraction with filtering
• Rich metadata extraction (YAML frontmatter + JSON)
• Reference guides for handling common extraction challenges

Sample prompt to use the skill:

<pathname>
Use "extracting-pdfs" skill to convert this pdf into a markdown file.

Source: Converted from anki-flashcard/prompt-v4.txt

Generate study flashcards from PDF or Markdown content in Anki-importable format. Card design follows evidence-based principles to optimize active recall and long-term retention under spaced repetition scheduling.

Trigger: Only when "Anki flashcard" or "Anki deck" is explicitly mentioned.

Process:

1. Read the source file (PDF or Markdown) thoroughly
2. Verify accuracy of all information in the source — correct any errors
3. Identify all key content: bolded terms, highlighted text, and Higher Tier material
4. Generate flashcards covering all essential topic content, selecting the most effective card type for each piece of knowledge
5. Run the interference check: scan the full card set for confusable pairs and add discriminative cards where needed
6. Format output as one card per line: Question | Answer

Core Design Principles:

• Understand-first rule: Never create cards for content the learner has not yet studied — cards consolidate existing understanding, they don't teach new concepts
• Minimum information (atomicity): Each card tests exactly one atomic piece of knowledge, answerable in under 6 seconds
• Production over recognition: Cards require producing an answer from memory, not merely recognizing it — no yes/no or true/false questions
• Depth of processing: Frame questions using "why," "how," or "explain" to force elaborative processing rather than rote retrieval
• Dual coding: Extract factual content from diagrams/visuals into text-based cards; the learner may attach images manually in Anki after import
• Personal connection: Where content allows, frame cards using concrete, relatable scenarios rather than abstract statements

Card Types:

• Definition (forward + reverse): Both directions for key terms to build bidirectional retrieval links
• Explain/justify: Full cause-and-effect chains for reasoning-based exam questions
• Cloze-style: Facts embedded in context with exactly one keyword deleted per card
• Compare/contrast: For commonly confused concepts — highlights the specific point of divergence
• Formula/equation: Formula recall plus at least one application card for when/why to use it
• Enumeration: Individual cards per list item rather than testing entire lists at once

Card Design Rules:

• Concise: Simple, direct language; aim for answers under 25 words for factual cards
• Unambiguous: Each question must have exactly one correct answer
• Bidirectional: Both forward and reverse cards for key definitions
• Interference management: Dedicated compare/contrast cards for confusable pairs
• Exclusions: No diagram-dependent questions, multi-step calculations, yes/no questions, or cards listing more than 3 items

Coverage Guidance: Prioritise content by examinability — definitions, laws, key equations, and explain/justify points first; supplementary context and edge cases only if the source emphasises them.

Output Format:

One card per line, question and answer separated by a pipe:

What is the unit of electrical resistance? | Ohm (Ω)
Define specific heat capacity | The energy required to raise the temperature of 1 kg of a substance by 1°C
The energy required to raise 1 kg of a substance by 1°C — what quantity is this? | Specific heat capacity
The SI unit of energy is the [...] | joule (J)
Explain why resistance increases with temperature in a metal | At higher temperatures, metal ions vibrate with greater amplitude, so conduction electrons collide more frequently with ions, transferring less charge per unit time
How does electrical conduction differ between metals and semiconductors? | In metals, resistance increases with temperature (more ion vibrations impede electron flow). In semiconductors, resistance decreases with temperature (more electrons gain enough energy to enter the conduction band, increasing the number of charge carriers)

Sample prompt to use the skill in claude.ai:

Use "anki-flashcard-generator" skill to create an Anki flashcard deck of the study materials.
Output the flashcards as a text file. Name the file after the source file (e.g., Physics_Chapter_5.pdf → Physics_Chapter_5.txt).

Sample user prompt to use the skill in Claude API:

Use "anki-flashcard-generator" skill to create an Anki flashcard deck of the study materials.
Output only the flashcard lines in the format "Question | Answer", one per line.
Do not include any preamble, headers, explanations, markdown formatting,
or code fences. The raw output will be saved directly to a text file.

Source: Converted from revision-notes/prompt-v2.txt

Generate concise, accurate revision notes from PDF or Markdown content.

Trigger: When asked to create revision notes, study notes, topic summaries, or condensed notes.

Process:

1. Read the source file thoroughly
2. Identify key content and Higher Tier material
3. Verify accuracy of all information
4. Write concise notes covering essential knowledge
5. Output as structured markdown file

Writing Guidelines:

• Concise: Condense to essential points
• Complete: Cover all necessary knowledge
• Accurate: Cross-check and correct errors
• Structured: Clear headings and logical organisation
• Higher Tier: Include and optionally mark with (HT)

Output Format: Markdown with title, section headings, bold key terms, and equations in code blocks.

Sample prompt to use the skill:

Use "revision-notes-generator" skill to create revision notes of the study materials.

Use "revision-notes-generator" skill to create revision notes of the study materials with title "<title>".

The following skills have been moved to the archive/ folder. They are preserved for reference but are no longer actively maintained.

Location: archive/pdf-extract/

The development/original version of the PDF extraction skill. Contains the core Python extraction script.

Note: For production use, see extracting-pdfs which is the current active version.

Features:

• Same extraction capabilities as extracting-pdfs
• Contains extract_pdf.py script (1,500+ lines)
• Full reference documentation included

Location: archive/pdf-to-markdown-converter/

The original PDF conversion skill using visual PDF understanding.

Status: Deprecated. Superseded by the more sophisticated extracting-pdfs skill.

pip install pymupdf pymupdf4llm

python extract_pdf.py <input_pdf> [output_folder] [options]

Options:

output_folder/
├── {filename}.md       # Markdown with YAML frontmatter
├── metadata.json       # Full extraction metadata
└── images/             # Extracted images

The PDF extraction skills include reference guides for handling common challenges:

• Extract everything — No hardcoded cleanup rules during extraction
• Preserve raw content — Keep data intact for intelligent post-processing
• Rich metadata — Provide comprehensive context for document understanding
• Manual over automated — Complex decisions handled manually for better results
• Atomic flashcards — One fact per card for effective learning
• Accuracy first — Verify and correct information in study materials

• Python 3 — Core scripting language
• PyMuPDF (fitz) — Low-level PDF reading and image extraction
• PyMuPDF4LLM — Enhanced markdown formatting with table support
• YAML/JSON — Metadata formats
• Markdown — Output format

This repository uses unified versioning. All skills share a single version number. See Releases for version history.

This work is licensed under CC BY 4.0 - you're free to share and adapt with attribution.