Draw.io Reconstruction: Editable Diagrams from Images

    1

    Reconstruct diagram images into editable Draw.io files with a Codex workflow, local helper scripts, and independent visual review.

    Free

    Security scanned

    Works with the AI Agents you already use

    Claude CodeClaude CodeCursorCursorCodex CLICodex CLIMuseMuseDotsDots+21 more

    About this skill

    Turn reference diagrams, research figures, architecture diagrams, and slide images into editable .drawio files. Native diagram elements preserve editable text and structure; complex visuals may remain embedded image crops. Includes independent producer/reviewer roles, visual comparison, and batch processing.

    The package contains skill instructions, Python helpers for cropping, validation and preview export, reference guides, and agent metadata. Outputs include Draw.io files, PNG previews and review audit records.

    Requires an agent with image inspection, local filesystem and shell access, and subagent support; Python 3.10+, Pillow, and Draw.io Desktop/CLI. Model usage is provided separately. Results depend on the model, input and available tools.

    Free MIT-licensed version by the original project author. Source and examples: https://github.com/sxy1499894281/drawio-reconstruction-skill

    Agensi MCP delivery compatibility is not yet verified. Supporting scripts and references must be available locally; if delivery provides instructions only, install the public repository locally.

    Requires Draw.io CLI for exports. Complex scene illustrations are handled as PNG crops to maintain fidelity. Independent review requires multi-agent support.

    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

    Free forever

    No account required to browse

    Trust & safety

    Security scan

    Security scan passed 18 days ago

    • Free with an Agensi account

    Requires

    Codex with subagent support
    Python 3.10+
    Pillow
    Draw.io Desktop/CLI
    Python

    Permissions

    Terminal / Shell
    Read Files
    Write Files
    Environment Variables

    Reads user-selected reference images; writes Draw.io files, crops, previews and audit files to the requested output folder. Runs local Python and Draw.io commands. May read DRAWIO_PATH to locate Draw.io. Uses independent subagents for visual review.

    Listed18 days ago

    What's inside

    Frequently Asked Questions

    The workflow separates production from independent visual review. It rebuilds text and structure with native Draw.io elements and checks exported previews against the reference. Visual fidelity depends on the source image, model and available tools; pixel-perfect results are not guaranteed.

    Outputs include editable .drawio XML, PNG previews and audit records. Text and structure use native elements where appropriate; complex visuals may remain embedded image crops. Optional PPTX conversion must be explicitly requested and requires suitable presentation tooling.

    Requires an agent with image inspection, local filesystem and shell access, and independent subagent support; Python 3.10+, Pillow for cropping, and Draw.io Desktop/CLI for preview export. Model usage is provided separately.

    The skill is designed to handle complex architectural diagrams and UI wireframes; however, highly detailed artistic illustrations or hand-drawn sketches may be processed as high-quality image crops rather than fully editable vector shapes.

    Yes. For two or more images, the workflow creates a batch manifest and queues per-image jobs within available agent capacity. Each image retains independent production and review stages. Concurrency depends on the runtime's available agent slots.

    New in Technical Documentation

    Free