ai-full-figure

Name: ai-full-figure
Author: neuromechanist

37stars7forks

How to install ai-full-figure?

neuromechanist/research-skills/ai-full-figure

$npx -y skills add neuromechanist/research-skills --skill ai-full-figure

Installs into the current project.

Summary

This skill should be used when the user asks to "generate an AI figure", "make a full AI-generated figure", "make a poster figure", "make a presentation figure", "generate a figure with AI", "create a substrate image", "AI background for a figure", "create a poster background", o

Files · 1

View on GitHub

SKILL.md

1	# AI Full Figure
2
3	Generate a pictorial substrate via AI (Codex CLI or OpenAI Images API) and overlay labels, arrows, and scale bars programmatically. The output is an SVG that embeds the raster substrate and lets the `[[scientific-figure]]` composer place it as a panel.
4
5	## When to use this skill
6
7	Reach for `ai-full-figure` when the figure's picture is the point:
8
9	- A brain rendering for a methods overview slide
10	- A microscope or apparatus illustration for a poster
11	- An anatomical scene as the background of a graphical abstract
12	- Any "what does this setup look like" pictorial
13
14	Reach for a different skill when the figure is information-dense:
15
16	\| Figure type \| Skill \|
17	\|---\|---\|
18	\| Data plot (matplotlib, seaborn, plotly) \| `[[plot-styling]]` for the plot, then `[[scientific-figure]]` to compose \|
19	\| Schematic with boxes / arrows / process flow \| `[[svg-primitives]]` (programmatic) or `[[svg-figure]]` (hand-authored conventions) \|
20	\| Flat scientific icon as part of a figure \| `[[transparent-icons]]` \|
21	\| Multi-panel journal figure \| `[[scientific-figure]]` as the composer \|
22
23	## Hard ceiling on AI generation
24
25	`gpt-image-2` and equivalent models cannot reliably render:
26
27	- Data plots with axis numerals (axes hallucinate, tick labels are nonsense)
28	- Equations (LaTeX or Greek-letter math)
29	- Multi-arrow flowcharts (arrow direction inconsistent, arrows cross wrong)
30	- Long labels in specific positions ("> 5 labeled elements" is a useful threshold)
31	- Anything that needs to be read precisely — labels >1–2 words drift
32
33	The right pattern when those are needed: substrate-only generation, programmatic overlay.
34
35	1. Prompt the model for the pictorial scene with no text, no labels, no arrows.
36	2. Compose labels, arrows, scale bars, and panel letters as a separate SVG layer.
37	3. Ship the combined SVG.
38
39	If the user is asking for a figure that violates the hard ceiling, route to `[[svg-figure]]` or `[[scientific-figure]]` and explain why. AI generation for embedded text remains unreliable as of 2026.
40
41	## The theme bible (shared with transparent-icons)
42
43	A `theme.json` keeps an AI-generated substrate visually consistent with the rest of the figure's palette and stroke language. The schema is the same one used by `[[transparent-icons]]`:
44
45	```
46	plugins/figures/skills/transparent-icons/references/theme.schema.json
47	```
48
49	Attach the theme to every generation call. The same fields apply: `palette`, `stroke`, `style_tokens`, `negative_tokens`, `composition`, `reference_images`, `model_preferences`. For substrates, the most load-bearing keys are:
50
51	- `palette.primary / accent / neutral` — color tokens injected into the prompt.
52	- `negative_tokens` — must include `text`, `labels`, `watermark`, `arrows`, `caption`. The substrate must not contain any text the overlay will replace.
53	- `composition.aspect` — substrates are typically 16:9 (`"1920x1080"`) or 4:3; set this explicitly.
54	- `style_tokens` — e.g. "photorealistic illustration", "watercolor", "isometric 3D", "schematic line art".
55	- `reference_images` (optional, up to 16) — for consistency across a set of substrates.
56
57	## Usage
58
59	### Generate a substrate
60
61	```bash
62	uv run --with openai --with python-dotenv --with pillow \
63	python scripts/generate_figure.py \
64	"a stylized lateral view of a human brain in soft watercolor on a clean white background" \
65	-o out/brain_substrate.png \
66	--size 1920x1080 --backend codex
67	```
68
69	`generate_figure.py` builds the prompt from the theme bible (when `--theme path/to/theme.json` is provided) and dispatches to the Codex CLI or the OpenAI Images API. Output is an opaque PNG at the requested size. No transparency post-process is applied — substrates are designed to be the bottom layer, not floated.
70
71	### Compose labels on top
72
73	```bash
74	uv run --with svgwrite --with pillow \
75	python scripts/overlay_labels.py \
76	out/brain_substrate.png \
77	--label "lateral sulcus@600,420" \
78	--label "central sulcus@800,300" \
79	--scale-bar "1 cm@200,950" \
80	-o out/brain_labeled.svg
81	```
82
83	The output SVG embeds the PNG via `<image>` and adds `<text>` / `<line>` / arrow markers on top. The SVG's `viewBox` matches the PNG's pixel dimensions so coordinates are pixel-address

People who liked this also viewed

6 picks

Type

ai-full-figure

37stars7forks

View on GitHub

How to install ai-full-figure?

neuromechanist/research-skills/ai-full-figure

$npx -y skills add neuromechanist/research-skills --skill ai-full-figure

Installs into the current project.

Summary

Files · 1

View on GitHub

SKILL.md

1	# AI Full Figure
2
3	Generate a pictorial substrate via AI (Codex CLI or OpenAI Images API) and overlay labels, arrows, and scale bars programmatically. The output is an SVG that embeds the raster substrate and lets the `[[scientific-figure]]` composer place it as a panel.
4
5	## When to use this skill
6
7	Reach for `ai-full-figure` when the figure's picture is the point:
8
9	- A brain rendering for a methods overview slide
10	- A microscope or apparatus illustration for a poster
11	- An anatomical scene as the background of a graphical abstract
12	- Any "what does this setup look like" pictorial
13
14	Reach for a different skill when the figure is information-dense:
15
16	\| Figure type \| Skill \|
17	\|---\|---\|
18	\| Data plot (matplotlib, seaborn, plotly) \| `[[plot-styling]]` for the plot, then `[[scientific-figure]]` to compose \|
19	\| Schematic with boxes / arrows / process flow \| `[[svg-primitives]]` (programmatic) or `[[svg-figure]]` (hand-authored conventions) \|
20	\| Flat scientific icon as part of a figure \| `[[transparent-icons]]` \|
21	\| Multi-panel journal figure \| `[[scientific-figure]]` as the composer \|
22
23	## Hard ceiling on AI generation
24
25	`gpt-image-2` and equivalent models cannot reliably render:
26
27	- Data plots with axis numerals (axes hallucinate, tick labels are nonsense)
28	- Equations (LaTeX or Greek-letter math)
29	- Multi-arrow flowcharts (arrow direction inconsistent, arrows cross wrong)
30	- Long labels in specific positions ("> 5 labeled elements" is a useful threshold)
31	- Anything that needs to be read precisely — labels >1–2 words drift
32
33	The right pattern when those are needed: substrate-only generation, programmatic overlay.
34
35	1. Prompt the model for the pictorial scene with no text, no labels, no arrows.
36	2. Compose labels, arrows, scale bars, and panel letters as a separate SVG layer.
37	3. Ship the combined SVG.
38
39	If the user is asking for a figure that violates the hard ceiling, route to `[[svg-figure]]` or `[[scientific-figure]]` and explain why. AI generation for embedded text remains unreliable as of 2026.
40
41	## The theme bible (shared with transparent-icons)
42
43	A `theme.json` keeps an AI-generated substrate visually consistent with the rest of the figure's palette and stroke language. The schema is the same one used by `[[transparent-icons]]`:
44
45	```
46	plugins/figures/skills/transparent-icons/references/theme.schema.json
47	```
48
49	Attach the theme to every generation call. The same fields apply: `palette`, `stroke`, `style_tokens`, `negative_tokens`, `composition`, `reference_images`, `model_preferences`. For substrates, the most load-bearing keys are:
50
51	- `palette.primary / accent / neutral` — color tokens injected into the prompt.
52	- `negative_tokens` — must include `text`, `labels`, `watermark`, `arrows`, `caption`. The substrate must not contain any text the overlay will replace.
53	- `composition.aspect` — substrates are typically 16:9 (`"1920x1080"`) or 4:3; set this explicitly.
54	- `style_tokens` — e.g. "photorealistic illustration", "watercolor", "isometric 3D", "schematic line art".
55	- `reference_images` (optional, up to 16) — for consistency across a set of substrates.
56
57	## Usage
58
59	### Generate a substrate
60
61	```bash
62	uv run --with openai --with python-dotenv --with pillow \
63	python scripts/generate_figure.py \
64	"a stylized lateral view of a human brain in soft watercolor on a clean white background" \
65	-o out/brain_substrate.png \
66	--size 1920x1080 --backend codex
67	```
68
69	`generate_figure.py` builds the prompt from the theme bible (when `--theme path/to/theme.json` is provided) and dispatches to the Codex CLI or the OpenAI Images API. Output is an opaque PNG at the requested size. No transparency post-process is applied — substrates are designed to be the bottom layer, not floated.
70
71	### Compose labels on top
72
73	```bash
74	uv run --with svgwrite --with pillow \
75	python scripts/overlay_labels.py \
76	out/brain_substrate.png \
77	--label "lateral sulcus@600,420" \
78	--label "central sulcus@800,300" \
79	--scale-bar "1 cm@200,950" \
80	-o out/brain_labeled.svg
81	```
82
83	The output SVG embeds the PNG via `<image>` and adds `<text>` / `<line>` / arrow markers on top. The SVG's `viewBox` matches the PNG's pixel dimensions so coordinates are pixel-address

People who liked this also viewed

6 picks

Type