declarative-engine-figures
Skill13 starsChanged 33 days agoMITView on GitHub
Grid-based block diagrams, network overview figures and deployment/phase flowcharts via a declarative layout engine — config syntax with real manuscript configs, and a no-download SVG extraction path.
In this repository8 other files it gives its agents.
- codex-figure-workflowSkill
- data-figures-rerenderSkill
- game-arch-diagramSkill
- hand-svg-structure-figureSkill
- paper-figureSkill
- parametric-diagram-scriptSkill
- style-presetsSkill
- visual-acceptance-loopSkill
The file
We split this skill into the rules it gives an agent
1 rule. About 0 are common, 0 are distinctive, and 1 is very specific to this project.
≈0
Common · estimate
Many files like this have a rule of this kind.
≈0
Distinctive · estimate
This author's own choices. Worth a look.
1
Very specific
Tied to their project's names, files or tools. Probably skip.
- Common
- Distinctive
- Very specific
- Very specific: line 23
Write the config; base64url-encode it into the hash: engine.html#c=<b64url>.
≈ “Common” is an estimate: at least 1 in 5 of the 350,495 skill files we read have a rule of the same kind, not the same words. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.
Files
Declarative layout-engine figures
When to use
Your figure fits a row×column grid: network overview figures (backbone/neck/head columns with legends and column heads), deployment or phase flowcharts (stacked dashed phase frames), method framework diagrams. A declarative engine does layout, alignment and orthogonal routing; you only write the configuration — and the same config always renders the same figure.
The configs in this folder were authored for a local companion engine (grid + emphasis colors + lane-based orthogonal routing + dashed groups + captions + true sub/superscripts). The engine itself is not bundled here; the config syntax and workflow below are the transferable part — port them to your engine of choice, or keep them as the spec.
Config syntax (real examples in configs/)
- Node:
name: main text | sub text @rRowcCol !blue— spans@r1-3c1(rows),@r2c3-4(columns);nat= natural-width narrow node; twonatnodes in one cell pack side-by-side. - Column heads
@head 1: Backbone; bottom legend@legend blue: P2 branch, orange: ACR; dashed phase group@group Training only : dashed … @end; bottom caption@caption …. - Edges
A -> B, coloredA -> B !orange; sub text supports_{sub}^{sup}(stackable:z^{T}_{c}) and\nforced line breaks. - Text arrows like
(32→64)inside node text are never mistaken for edges (position-based disambiguation).
Workflow (no-download SVG extraction)
- Write the config; base64url-encode it into the hash:
engine.html#c=<b64url>; success =window.__archState.ok === true. - Evaluate
buildSVG(lastScene)in the page console — it returns the full SVG string (no browser download handling needed). - The exported width/height are float strings — later regexes must accept
[\d.]+. - Export PNG via the
hand-svg-structureinline-page pipeline.
Real configs (configs/)
fig4-1.config.txt— network overview with column heads + legend + packed narrow-node pair.fig5-2.config.txt— teacher–student distillation training framework.fig6-2_convert.config.md— TensorRT deployment flow (two stacked dashed phase frames).
Does it work?
Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.
No reviews yet
People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.
- It worked
- Partly
- It didn't
Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.
Your trail
Files you open will collect here, so you can find them again.
Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.
Now do this for your own.
Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.

