More screenshots

    Works with the AI Agents you already use

    Claude CodeClaude CodeCursorCursorCodex CLICodex CLIMuseMuseDotsDots+21 more

    research diagram architect

    1

    Turn research pipelines and system architectures into editable PowerPoint, TikZ, Mermaid, and SVG diagrams from one structured specification.

    Free

    1 useSecurity scanned
    research diagram architect

    research diagram architect

    Example session with this skill installed

    Create an academic diagram illustrating a Hybrid RAG architecture.

    Separate the workflow into two phases

    1. Offline Indexing: Raw Documents → Chunking → Embedding Model → Vector Database.
    2. Online Retrieval: User Query → Query Embedding → Top-K Retrieval → Context Assembly → LLM → Final Answer.

    Connect the Vector Database to Top-K Retrieval with a dashed reference arrow.

    Use the academic minimalist theme, emphasize key components, and export editable PowerPoint, TikZ, Mermaid, and SVG files.

    Validate the diagram for node collisions and text overflow.

    • Read your context and instructions
    • Compiled the research diagram architect

    Hybrid RAG Architecture — Generated Diagram

    Created a structured Hybrid Retrieval-Augmented Generation diagram from a single semantic specification.

    Architecture

    Offline Indexing

    Raw Documents → Chunking → Embedding Model → Vector Database

    Online Retrieval

    User Query → Query Embedding → Top-K Retrieval → Context Assembly → LLM Generation → Final Answer

    A dashed connection links the Vector Database to Top-K Retrieval, representing the cross-phase similarity search operation.

    Generated Outputs

    PowerPoint (.pptx)
    An editable technical diagram using native shapes and connectors, suitable for modifying directly in presentation software.

    LaTeX TikZ (.tex)
    Diagram source code for integration into research papers and LaTeX projects.

    Mermaid (.mmd)
    A text-based architecture diagram that can be embedded into GitHub README files or Markdown documentation.

    SVG (.svg)
    A scalable vector diagram for technical documentation, websites, and research presentations.

    Design & Validation

    The diagram uses the academic minimalist preset with neutral colors, a consistent layout, and highlighted key components.

    The semantic specification defines each node and connection independently of the export format, allowing the same architecture to be reused across multiple targets.

    Geometric validation checks node positioning, collisions, and estimated text overflow before delivery.

    Result: One reusable architecture specification with four export formats, reducing the need to redraw the same technical figure for different purposes.

    PowerPoint (.pptx

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

    About this skill

    Research Diagram Architect

    One specification. Four editable diagram formats.

    Research Diagram Architect helps researchers, developers, and students create structured, publication-ready technical diagrams without manually rebuilding the same figure for different platforms.

    Whether you're documenting an AI pipeline, designing a system architecture, or preparing a research paper, this skill provides a reusable workflow from semantic specifications to professional diagrams.

    What You Can Create

    • AI & ML Pipelines: RAG systems, neural networks, retrieval workflows, and inference architectures.
    • System Architectures: Cloud services, databases, APIs, and distributed systems.
    • Research Workflows: Experiment pipelines, data processing, and methodology diagrams.
    • Technical Documentation: Clear, structured diagrams for papers, presentations, and project documentation.

    Four Output Formats

    • PowerPoint (.pptx): Editable vector shapes for presentations.
    • LaTeX TikZ (.tex): Diagram source code for academic papers.
    • Mermaid (.mmd): Diagram code for Markdown and GitHub.
    • SVG (.svg): Scalable graphics for web and documentation.

    Built for Research

    • Semantic nodes, containers, and connections.
    • Four predefined visual themes.
    • Deterministic geometry and text-overflow checks.
    • Reusable specifications for consistent exports.

    Getting Started

    Use the skill to structure your diagram, then render it using the open-source Python package. Natural-language planning requires a configured compatible LLM provider.

    Open source: https://github.com/quangnhat1504/research-diagram-architect

    Build diagrams once. Refine and export them wherever your research takes you.

    Requires external Python package installation. Output quality for complex overlapping containers depends on the compute_layout engine. Visual QA requires a compatible rendering environment.

    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

    1 use

    Times used to date

    Free forever

    No account required to browse

    Trust & safety

    Security scan

    Security scan passed 1 month ago

    • Free with an Agensi account

    Requires

    Python

    Permissions

    No special permissions declared or detected

    Listed1 month ago
    Updated1 month ago

    What's inside

    Frequently Asked Questions

    Unlike standard AI images or flat exports, this skill generates editable files like PPTX shapes and TikZ code, allowing you to manually tweak nodes or change text directly within your presentation or LaTeX editor.

    The skill requires a Python environment with the `research-diagram-architect` package; it is designed to integrate with AI agents capable of executing Python code to perform the layout calculations and file generation.

    Yes, the skill uses a 'UniversalDiagramSpec' that allows you to export the exact same architecture design into Mermaid for GitHub, TikZ for your paper, and PowerPoint for your presentation simultaneously.

    Your purchase includes the core skill logic, the `UniversalDiagramSpec` schema, and access to the specialized layout engine drivers for all supported formats (PPTX, TikZ, Mermaid, and SVG).

    The skill includes a 'geometry linting' step that programmatically checks for overlapping nodes or text overflows, ensuring your diagrams are professional and readable before you download them.

    New in Technical Documentation