yizhiyanhua-ai/fireworks-tech-graph

• **Archetype**: utility-with-skill
• **Stars**: 1,512
• **Last pushed**: 2026-04-12
• **License**: MIT
• **Discovered**: 2026-04-12
• **Skills found**: 1

Summary

Claude Code skill for generating production-quality SVG+PNG technical diagrams. Supports 8 diagram types and 5 visual styles with deep AI/Agent domain knowledge. Uses rsvg-convert for SVG-to-PNG export. Includes 4 helper scripts: generate-diagram.sh (validate+export), generate-from-template.py (template-based SVG creation), validate-svg.sh (syntax validation), and test-all-styles.sh (batch testing). Bilingual triggers (Chinese + English).

Assessment

MEDIUM VALUE. The helper script architecture is well-designed: a Python template renderer produces SVGs from structured JSON input (title, nodes, arrows, legend), a bash validator checks XML syntax/tag balance/marker references/path data, and a test script batch-validates all styles. This validation-centric approach to SVG generation (generate -> validate -> export) is a reusable pattern for any code-generation skill. The template-based generation avoids the fragility of raw SVG string construction. The bilingual trigger pattern shows how to make skills accessible across language communities.

Extraction Priority

• Low-Medium
• Rationale: The diagram generation procedure and validation pipeline are useful patterns, but the skill is primarily a rendering tool. The template-based generation + validation loop is the most transferable element.

Processes

• **Validated Diagram Generation Process**: Select diagram type + style -> generate from template with structured JSON -> validate SVG (XML syntax, tag balance, markers, paths, attributes) -> export PNG. The validation step prevents malformed output.
• **Batch Style Testing**: Generate diagrams across multiple sizes and styles -> validate all -> produce test report. A CI-like quality gate for visual output.

Plugin Ideas

• **Tech Diagram Generator plugin**: Install.md-driven plugin that adds diagram generation skills with template-based SVG creation and validation. Configurable diagram types and visual styles.

Patterns

• **Template-based generation with validation**: Generate from structured input (JSON) via templates rather than raw string construction. Validate output before delivery. Prevents fragile code generation.
• **Helper script decomposition**: Four focused scripts (generate, template-render, validate, batch-test) each handling one concern. The skill orchestrates them as a pipeline.
• **Bilingual triggers**: Both Chinese and English trigger phrases in the skill description. Simple internationalization for skill discovery.
• **SVG validation checklist**: XML syntax, tag balance, marker references, attribute completeness, path data integrity. A reusable validation pattern for any structured output format.

Library Mapping

Plugin Marketplace Mapping