figure-spec
Generate deterministic publication-quality architecture, workflow, and pipeline diagrams from structured JSON (FigureSpec) into editable SVG. Use when user says "架构图", "workflow 图", "pipeline 图", "确定性矢量图", "figure spec", "draw architecture", or needs precise, editable, publication-ready vector diagrams. Preferred over AI illustration for formal architecture/workflow figures.
71
88%
Does it follow best practices?
Impact
—
No eval scenarios have been run
Passed
No known issues
Quality
Discovery
100%Based on the skill's description, can an agent find and select it at the right time? Clear, specific descriptions lead to better discovery.
This is an excellent skill description that clearly communicates specific capabilities (deterministic SVG diagram generation from FigureSpec JSON), provides rich bilingual trigger terms, explicitly states when to use it, and distinguishes itself from related skills like AI illustration. The description is concise yet comprehensive, covering input/output formats, diagram types, and disambiguation guidance.
| Dimension | Reasoning | Score |
|---|---|---|
Specificity | Lists multiple specific concrete actions: 'Generate deterministic publication-quality architecture, workflow, and pipeline diagrams from structured JSON (FigureSpec) into editable SVG.' This clearly describes the input format (JSON/FigureSpec), output format (editable SVG), and the types of diagrams supported. | 3 / 3 |
Completeness | Clearly answers both 'what' (generate deterministic publication-quality diagrams from FigureSpec JSON into editable SVG) and 'when' (explicit 'Use when' clause with specific trigger phrases, plus a disambiguation note about preferring this over AI illustration for formal figures). | 3 / 3 |
Trigger Term Quality | Excellent coverage of natural trigger terms including bilingual keywords ('架构图', 'workflow 图', 'pipeline 图', '确定性矢量图'), English terms ('figure spec', 'draw architecture'), and conceptual triggers ('publication-ready vector diagrams', 'editable'). Covers both Chinese and English user queries. | 3 / 3 |
Distinctiveness Conflict Risk | Highly distinctive with a clear niche: deterministic, structured JSON-based diagram generation for formal architecture/workflow figures. The explicit contrast ('Preferred over AI illustration for formal architecture/workflow figures') helps disambiguate from general illustration or freeform drawing skills. | 3 / 3 |
Total | 12 / 12 Passed |
Implementation
77%Reviews the quality of instructions and guidance provided to agents. Good implementation is clear, handles edge cases, and produces reliable results.
This is a well-structured, highly actionable skill with clear workflows and validation checkpoints. Its main weakness is verbosity—particularly the lengthy tool resolution bash block and some sections that could be offloaded to reference files. The concrete JSON templates and executable commands make it immediately usable, but the token cost is higher than necessary.
Suggestions
Move the Tool Location bash resolution chain to a separate reference file (e.g., TOOL_RESOLUTION.md) and summarize it in one line in the main skill body.
Move the Schema Quick Reference table to a separate SCHEMA.md file, keeping only a one-liner pointer in the main skill (the renderer already has `schema` subcommand).
| Dimension | Reasoning | Score |
|---|---|---|
Conciseness | The skill contains useful content but is verbose in several areas: the Tool Location section with its lengthy bash resolution chain is overly detailed, the 'When to Use' / 'Do NOT use' section explains routing decisions Claude could infer, and the optional Codex review step adds significant length for an edge case. The schema reference and design patterns are reasonably tight. | 2 / 3 |
Actionability | The skill provides fully executable bash commands, concrete JSON templates for multiple diagram types, specific render/validate/convert commands, and a complete schema reference table. The templates are copy-paste ready and cover the main use cases (architecture, workflow, decision diamond). | 3 / 3 |
Workflow Clarity | The 5-step workflow is clearly sequenced with explicit validation checkpoints: validate before render (Step 3), visual review checklist (Step 4) with the rule to edit JSON not SVG, and an optional cross-model review loop with scoring thresholds. The feedback loop for validation failures is explicitly stated. | 3 / 3 |
Progressive Disclosure | The skill references external files (shared-references/integration-contract.md, shared-references/review-tracing.md) appropriately, but the main SKILL.md itself is quite long (~200+ lines) with the Tool Location bash block and schema reference inlined rather than split to separate files. The schema quick reference and design patterns could be in referenced files to keep the main skill leaner. | 2 / 3 |
Total | 10 / 12 Passed |
Validation
81%Checks the skill against the spec for correct structure and formatting. All validation checks must pass before discovery and implementation can be scored.
Validation — 9 / 11 Passed
Validation for skill structure
| Criteria | Description | Result |
|---|---|---|
allowed_tools_field | 'allowed-tools' contains unusual tool name(s) | Warning |
frontmatter_unknown_keys | Unknown frontmatter key(s) found; consider removing or moving to metadata | Warning |
Total | 9 / 11 Passed | |
- Commit
66b974e
- Reviewed
Table of Contents
Is this your skill?
If you maintain this skill, you can claim it as your own. Once claimed, you can manage eval scenarios, bundle related skills, attach documentation or rules, and ensure cross-agent compatibility.