Works with the AI Agents you already use

    Claude CodeClaude CodeCursorCursorCodex CLICodex CLIMuseMuseDotsDots+21 more

    Flashcard Architect: PDF, DOCX & MD to Anki Deck Generator

    1

    Convert PDFs, DOCX, and Markdown into FSRS-optimized Anki decks (.apkg) with Y0 geometric diagram anchoring, SHA-256 WebP media, and automated mock exam auditing.

    $3.99

    Secure checkout via Stripe

    Money-back guarantee

    Converts to your local currency at checkout

    Security scanned
    Flashcard Architect: PDF, DOCX & MD to Anki Deck Generator

    Flashcard Architect: PDF, DOCX & MD to Anki Deck Generator

    Example session with this skill installed

    Convert my graded AWS_Solutions_Architect_Mock_Exam.pdf into an Anki deck named AWS::SAA-C03::Error_Review. Extract all VPC and architecture diagrams in their exact question positions, generate an exam audit report, and include Cloze deletions for key AWS service limits.

    • Read your context and instructions
    • Compiled the flashcard architect
    • Generated the document

    Flashcard Architect v3.2.0 — Execution Summary

    I have processed AWS_Solutions_Architect_Mock_Exam.pdf (42 pages, 65 questions) using the decoupled

    Universal Multi-Agent Workflow (--extract-only followed by --no-ai packaging).

    Document Parsing & Y0 Anchoring: Extracted 65 questions and 18 vector/bitmap architecture diagrams via PyMuPDF vertical bounding-box coordinates (bbox.y0).
    *

    Media Normalization: Converted all 18 diagrams into lossless/high-efficiency WebP format using SHA-256 content hashing (ag_4f9a2c1e.webp ... ag_9d81b40f.webp).
    *

    Exam Audit & Filtering: Identified 51 correct answers (status::correct) and 14 incorrect answers (status::wrong). Generated both the full deck and a targeted review report.


    Automated Exam Audit Report (AWS_SAA_C03_Audit_Report.md)

    Global IndexSource LocationTopic / DomainAudit StatusAnchored Media AssetCard Type
    Global #04Page 3, Q4VPC Peering & Transit Gatewaystatus::wrongag_4f9a2c1e.webp (Y0: 412.5)Basic + Diagram
    Global #12Page 8, Q12S3 Object Lock & Glacier Vaultstatus::wrongNoneCloze ({{c1::...}})
    Global #19Page 14, Q19Multi-AZ RDS vs Read Replicasstatus::correctag_7c3b910a.webp (Y0: 285.0)Basic + Diagram
    Global #27Page 21, Q27ALB Path-Based Routing & WAFstatus::wrongag_8e12d54b.webp (Y0: 518.2)Basic + Diagram
    Global #33Page 26, Q33DynamoDB Global Tables RCU/WCUstatus::wrongNoneCloze ({{c1::...}})

    Sample Generated Flashcards (Preview)

    [
      {
        "concept_id": "AWS-SAA-P03-Q04-G004",
        "question": "[Page 3 | Question 4 (Global #4)] A company connects 15 VPCs across 3 regions and needs centralized inspection for east-west traffic without full-mesh peering. Based on the topology below, which architecture minimizes operational overhead?<br><img src=\"ag_4f9a2c1e.webp\">",
        "answer": "<div class=\"ans\">AWS Transit Gateway with a centralized Inspection VPC and inter-region peering attachments.</div><br><b>Why Option B is incorrect:</b> VPC Peering does not support transitive routing, requiring 105 individual peering connections for 15 VPCs.",
        "tags": ["aws::networking", "status::wrong", "page::03"]
      },
      {
        "concept_id": "AWS-SAA-P08-Q12-G012",
        "question": "[Page 8 | Question 12 (Global #12)] To enforce SEC Rule 17a-4 WORM compliance where not even the AWS root account can delete objects until retention expires, enable S3 Object Lock in {{c1::Compliance mode}} rather than {{c2::Governance mode}}.",
        "answer": "Compliance mode locks retention settings permanently for the duration of the retention period, whereas Governance mode allows privileged IAM users with `s3:BypassGovernanceRetention` to override lock settings.",
        "tags": ["aws::storage", "status::wrong", "page::08"]
      }
    ]
    

    AWS_Solutions_Architect_Mock_Exam.pdf

    PDF · document

    Generated

    Example file from a real run - the skill writes it into your workspace.

    Connects securely to your tools. The creator never sees your data.

    About this skill

    Turn Dense Study Material & Exam Dumps into High-Retention Anki Decks

    Flashcard Architect (v3.2.0) is an end-to-end document parsing and spaced-repetition engineering skill designed for software engineers, IT certification candidates, medical students, and researchers. Instead of manually copying questions and screenshotting diagrams, your AI agent extracts text and visual assets locally and compiles a ready-to-import .apkg Anki package.


    Core Technical Capabilities

    • Geometric Y0 Diagram Anchoring (bbox.y0): Extracts vector graphics and bitmap diagrams from PDFs using PyMuPDF, reads their exact vertical page coordinates, and anchors each topology diagram, table, or chart directly beneath its corresponding question—eliminating floating or misplaced media.
    • Collision-Free SHA-256 WebP Media: Automatically converts and optimizes extracted images into lightweight WebP files named by content hash (ag_<sha256>.webp), ensuring Anki's global collection.media folder never overwrites diagrams across different decks.
    • Global Sequential Indexing: Tags and prefixes multi-part exam dumps with [Page X | Question Y (Global #Z)] to prevent card collisions across large question banks.
    • Automated Mock Exam Auditing (status::wrong / status::correct): Detects answer keys and user performance in graded practice tests, tags cards by accuracy, supports error-only deck generation (--filter wrong), and outputs an analytical Markdown report (_Audit_Report.md).
    • FSRS-Ready Dual Note Routing & Native Dark Mode CSS: Automatically routes standard Q&A pairs to Basic cards and {{c1::...}} deletions to Cloze cards, styled with a high-contrast dark slate and neon-green stylesheet (ANKI_CSS) built for rapid visual scanning.
    • Universal Multi-Agent Architecture: Built for Claude Code, Cursor, GitHub Copilot, Windsurf, and Gemini CLI via decoupled extraction (--extract-only) and deterministic packaging (--no-ai).

    1. Scanned PDFs composed purely of flat page scans without an embedded text layer require OCR preprocessing before text extraction.
    2. Requires Python 3.9+ installed on the local machine along with PyMuPDF, genanki, Pillow, and python-docx (can be installed in a virtual environment or via the explicit --auto-install flag).
    3. Bilingual JSON schema supports both English (question, answer, concept_id) and Portuguese (pergunta, resposta, id_conceito) keys natively.

    How to use

    1. 1

      Connect your agent to the Agensi MCP

      Copy the prompt and paste it into Claude, ChatGPT, Cursor or any agent that supports MCP. It sets up the connection for you. Manual setup

    2. 2

      Sign in to Agensi

      Your agent shows a sign-in link. Sign in and click Allow.

    3. 3

      Ask for what you need

      Your agent loads this skill and uses it on your task. It always asks before any payment.

    Reviews

    No reviews yet

    Be one of the first to try it. Every listed skill goes through the trust checks below.

    Security scanned

    Checked by our 8-point scan before listing

    Fresh listing

    Recently published to Agensi

    Money-back guarantee

    Doesn't do what its page promises? You get your money back.

    Trust & safety

    Security scan

    Security scan passed 14 days ago

    Requires

    Ankiweb
    Python 3.9+
    PyMuPDF (fitz)
    genanki
    Pillow
    python-docx
    Anki Desktop or AnkiMobile or AnkiDroid
    Python

    Permissions

    Terminal / Shell
    Read Files
    Write Files

    Allowed Hosts

    apps.ankiweb.net

    File Scopes

    **/*.pdf
    **/*.docx
    **/*.md
    **/*.txt
    **/*.json
    **/*.apkg
    **/*.webp

    Runs 100% locally via Python to parse PDFs/DOCX/Markdown, extract diagrams into SHA-256 hashed WebP files, and compile offline Anki (.apkg) packages. Complies with strict security standards: no global pip modifications and no elevated shell flags required.

    Listed14 days ago

    What's inside

    Frequently Asked Questions

    No. In Universal Multi-Agent mode (--extract-only and --no-ai), the Python script extracts the text and Y0-anchored diagrams locally, your current agent (Claude Code, Cursor, or Copilot) structures the flashcards using your existing session, and the script packages the final .apkg file offline.

    Yes. The engine automatically audits graded mock exams, applies status::wrong and status::correct tags, generates a full Markdown Audit Report, and supports the --filter wrong CLI flag to export decks containing only your mistakes.

    New in Education & Learning