CtrlK
BlogDocsLog inGet started
Tessl Logo
SkillPublic

baoyu-design

Create or edit visual design artifacts as self-contained HTML: UI mockups, interactive prototypes, wireframes, landing pages, dashboards, app screens, slide decks, print documents, résumés, animations, 3D objects, diagrams, fliers, HTML email, and visual research reports. Also author reusable design systems, UI kits, brand tokens, and component libraries, or export slide-structured decks made with this skill to PPTX. Use only when the task actually creates or edits a visual deliverable, even if the user does not say "design". Architecture or protocol analysis, rendering-engine or Rust/WASM/backend implementation, and technical Markdown do not trigger it by themselves. In a larger task, load it when visual creation or editing begins.

Quality

84%

Does it follow best practices?

Run evals on this skill

Adds up to 20 points to the overall score

View guide
SecuritybySnyk

Failed to scan

The risk profile of this skill

SKILL.md
Quality
Evals
Security

Quality

Content

73%Weight 40%Scale 1-5

Reviews the quality of instructions and guidance provided to agents. Good implementation is clear, handles edge cases, and produces reliable results.

The body is an effective routing document: a clear numbered entry flow with explicit verification, concrete paths/commands/tool names, and disciplined progressive disclosure of a large design system. Its main weakness is redundancy — the design-system binding instruction and the project-type routing are each repeated — and a few deferred details (script invocation syntax) that keep individual flows from being fully self-contained.

Suggestions

State the bound-design-system loading rule (read `_ds/<slug>/_ds_prompt.md` and follow it as a binding style) exactly once — in step 5 — and have the use-design-system bullet and resume note reference it instead of repeating it nearly verbatim.

Rely on the `project-types.json` routing table alone for the 13 canonical types and keep only the bullets that add non-table conditions (Figma/GitHub/HTML imports, design-system authoring, PPTX export policy, the something-cool special case), cutting the duplicated per-type bullets.

Include a one-line example invocation for the `agents/*.mjs` scripts (e.g. how `compile-design-system.mjs` is run) in the body so the design-system and Figma flows are executable without first consulting a harness reference doc.

DimensionReasoningScore

Conciseness

The body is mostly efficient and explains nothing Claude already knows, but it could be tightened: the instruction to load a bound design system's prompt ("read `_ds/<slug>/_ds_prompt.md`... follow it as a binding visual style") is repeated nearly verbatim three times (in the use-design-system bullet, step 5, and the resume note), step-3 bullets run 100+ words with dense parentheticals, and the 13-type routing table in `project-types.json` is restated as individual bullets for the same types. This matches anchor 3 ("mostly efficient but... could be tightened"); it is not 2 because there is no concept padding or filler explanation.

3 / 5

Actionability

Guidance is concrete throughout: exact file paths to read per condition, specific tool names (AskUserQuestion, SendUserFile, `functions.*`), a copy-paste glob pattern (`glob designs/*/_ds_manifest.json`), default output folder, and named scripts (`agents/compile-design-system.mjs`, `agents/check-design-system.mjs`, `agents/build-preview.mjs`, `agents/record-asset.mjs`) with subcommand names (`outline`, then `mount`/`materialize`/`render`). It stops short of anchor 5 because actual invocation syntax for the `agents/*.mjs` scripts is deferred ("see your harness reference for how to launch it"), leaving minor execution gaps in the design-system and Figma flows.

4 / 5

Workflow Clarity

The entry flow is a clearly sequenced 6-step process (load methodology → identify harness → load built-in skills → ask clarifying questions → set up output folder → build/preview/verify) with an explicit validation checkpoint and error-recovery loop in step 6 ("preview it over HTTP... and confirm it loads cleanly. Fix any errors before finishing") plus a state-recovery path for resumed projects (read `_d_meta.json` first). This matches anchor 5; it is not 4 because validation and the fix-before-finishing loop are explicit rather than implied, and step 4 front-loads a question checkpoint before any building.

5 / 5

Progressive Disclosure

SKILL.md is a well-signaled orchestrator overview pointing to one-level-deep references, and the cited harness docs (`references/claude.md`, `references/cursor.md`, `references/codex.md`, `references/upstream-system-prompt.md`, `references/upstream-sync/`) all exist in the bundle as linked. It falls short of anchor 5: the core `system-prompt.md`, `project-types.json`, `built-in-skills/`, `agents/`, and `starter-components/` files referenced throughout are not present in the provided bundle, and some routing chains nest two levels (e.g. the design-system bullet routes to `design-system-authoring-guide.md`, which itself routes to `create-design-system.md`, `design-components.md`, and the checker scripts), which is a minor organization/navigation gap.

4 / 5

Total

16

/

20

Passed

Description

88%Weight 40%Scale 1-5

Based on the skill's description, can an agent find and select it at the right time? Clear, specific descriptions lead to better discovery.

A strong description: concrete, comprehensive, third-person, with an explicit and unusually precise when-to-use clause including negative boundaries. The only weakness is trigger-term coverage that stops just short of comprehensive — a few common synonyms and file extensions are missing.

DimensionReasoningScore

Specificity

The description enumerates multiple concrete actions ("Create or edit visual design artifacts as self-contained HTML", "author reusable design systems, UI kits, brand tokens, and component libraries", "export slide-structured decks... to PPTX") across a comprehensive list of 15+ deliverable types (UI mockups, wireframes, dashboards, slide decks, résumés, 3D objects, fliers, HTML email). Coverage is comprehensive with no obvious gaps, matching the anchor-5 example's breadth; it is not 4 because the action list and type enumeration leave no minor coverage gaps.

5 / 5

Completeness

It explicitly answers both questions: the "what" is the concrete creation/editing/authoring/export list, and the "when" is fully explicit with concrete trigger guidance — "Use only when the task actually creates or edits a visual deliverable, even if the user does not say 'design'" and "In a larger task, load it when visual creation or editing begins", plus explicit non-triggers. This matches anchor 5; it exceeds anchor 4 because the when-clause is specific and unambiguous rather than merely present.

5 / 5

Trigger Term Quality

Strong natural keyword coverage ("UI mockups", "wireframes", "landing pages", "dashboards", "slide decks", "résumés", "fliers", "HTML email", "design systems", "PPTX") that directly matches what users would say. It falls between anchors 4 and 5: a few common natural terms and synonyms are missing ("presentation", "poster", "infographic", "website") and file extensions beyond PPTX (.html, .fig) are absent, so it does not reach the comprehensive synonym-plus-extension coverage of anchor 5.

4 / 5

Distinctiveness Conflict Risk

The skill has a clear niche (self-contained HTML visual deliverables) and reduces conflict with explicit exclusions ("Architecture or protocol analysis, rendering-engine or Rust/WASM/backend implementation, and technical Markdown do not trigger it by themselves"). It is not 5 because "landing pages", "dashboards", and "diagrams" carry genuine overlap risk with web-development and data-visualization skills, keeping it at anchor 4 (mostly distinct, minor overlap with closely related skills).

4 / 5

Total

18

/

20

Passed

Validation

93%

Checks the skill against the spec for correct structure and formatting. All validation checks must pass before discovery and implementation can be scored.

Validation — 15 / 16 Passed

Validation for skill structure

CriteriaDescriptionResult

relative_links

Relative link issues: 25 missing

Warning

Total

15

/

16

Passed

Repository

JimLiu/baoyu-design

Path

archive/baoyu-design/SKILL.md

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.