- Home
- Skills
- Technical Documentation
- figure designer
Works with the AI Agents you already use
figure designer
Turn any methodology, pipeline, or result into a publication-ready figure, with spec, caption, and renderable code.
$7
figure designer
Example session with this skill installed
Use figure-designer to create the methodology overview figure for my paper on a transformer-based time-series forecasting model with a decomposition block.
- Read your context and instructions
- Compiled the figure designer
Designed a single-column methodology overview figure with 6 modules, 8 labeled arrows, a color-blind-safe palette, a ready-to-render Mermaid diagram, and a numbered caption.
Figure brief
Message: the model decomposes the series, then forecasts each component with a transformer and recombines the outputs.
Visual specification
| Element | Type | Label | Notes |
|---|---|---|---|
| Input series | Node | x_t | Historical time series |
| Decomposition block | Node | Trend + seasonal | Splits input |
| Transformer encoder | Node | Encoder | Processes each component |
| Forecast head | Node | Linear head | Produces per-component forecast |
| Recombination | Node | Sum | Merges components |
| Output | Node | y_hat | Final forecast |
Renderable code (Mermaid)
flowchart LR
A[Input x_t] --> B[Decomposition]
B --> C[Trend]
B --> D[Seasonal]
C --> E[Encoder]
D --> E
E --> F[Forecast head]
F --> G[Recombine]
G --> H[Output y_hat]
Caption
Figure 1. Overview of the proposed model. The input series is decomposed into trend and seasonal components, each encoded by a shared transformer, and the forecasts are recombined into the final output.
Next steps
- Run
paper-outlineto confirm the figure's page placement. - Run
reviewerto stress-test the figure clarity before submission.
Connects securely to your tools. The creator never sees your data.
About this skill
The problem
Reviewers skim figures before reading the text, and a confusing diagram can sink an otherwise strong paper. Researchers waste hours in drawing tools and still produce figures that are unclear, inconsistent, or illegible at print size. This skill turns the method and results into publication-ready figures.
What it does
- Extracts components, data flow, and the training/inference pipeline from your methodology.
- Chooses the right diagram type, such as overview, architecture, pipeline, flowchart, comparison, or ablation.
- Produces a precise visual specification with boxes, arrows, labels, and a color-blind-safe palette.
- Generates renderable code in Mermaid, Graphviz DOT, TikZ, draw.io XML, or matplotlib.
- Writes the numbered caption, alt text, and in-text reference wording.
- Checks venue compliance for column width, font size, and grayscale fallback.
Why this beats prompting it yourself
Generic prompts describe the figure but leave the layout to you. This skill extracts the single message the figure must convey, composes the diagram element by element, and emits code that actually renders, so you can drop it straight into Overleaf, draw.io, or a README.
Use cases
- Designing the methodology overview or architecture figure (Fig. 1).
- Turning a written pipeline into a block diagram or flowchart.
- Creating results figures such as comparison charts, ablation heatmaps, and confusion matrices.
- Writing captions and alt text that survive a skim read.
- Making existing figures venue-compliant and color-blind safe.
Known limitations
The skill produces a specification and renderable code, not a raster image; the code must be rendered in a compatible viewer. The quality of a data figure depends on the accuracy of the numbers you provide.
Important
It designs and specifies figures; it does not run experiments or generate results data. Feed it the methodology from methodology-designer and the figure list from paper-outline for the best output.
Why buy it
Use it after methodology-designer and paper-outline to turn the written method into clear, publication-ready visuals. It is the figure step of the design pack.
Generates code and specs only; does not produce raster images. Requires external rendering (Overleaf, Python, or Draw.io). Accuracy of data figures depends on user-provided numbers.
How to use
- 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
Sign in to Agensi
Your agent shows a sign-in link. Sign in and click Allow.
- 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 1 month ago
Requires
Permissions
No special permissions declared or detected