Create research diagrams with AI
Turn a research workflow or argument into editable Mermaid source, keeping the figure’s question, evidence and relationship labels visible to reviewers.
To create research diagrams with AI, begin with the question the figure needs to answer. A selection flow can explain how studies moved between stages; a conceptual model may need to distinguish proposed links from observed relationships. The research-diagrams skill chooses a form according to that purpose.
A diagram can claim too much without using a single sentence. An unlabelled arrow may suggest causation when the evidence only establishes association. The workflow asks for a caption and an evidence record so that the meaning of the connections is clear. Visual validation remains a separate step from checking the source text.
What it does
- Map stages in a documented research process.
- Separate decision points from ordinary sequence steps.
- Label observed, conceptual and proposed relationships appropriately.
- Prepare editable Mermaid source and a caption for the draft.
- Check rendered labels, reading order and the relationship to the evidence.
Install and use
- Install the skill with its research workspace guidance in a private project. Provide the relevant version and source records. Use a Mermaid-compatible renderer to verify the appearance.
- Ask, “Diagram our screening workflow from these notes. Keep planned steps distinct and flag counts that have not been verified.”
- Read the caption and figure together. Check the rendered output before treating the visual review as complete.
FAQ
Source
Made by Anh Dang. Original skill: research-diagrams, published on GitHub under the MIT licence.
View on GitHub