ljg-present

Unix 气质的演讲设计与保真排版。把 Org/Markdown 提纲或经授权提炼的讲稿制成单文件离线 HTML；以语义角色选择巨句、对照、Unicode 关系图、证据与章节版式。默认深色，支持讲稿层、原生图表与翻页笔。USE WHEN present、做演讲、slides、按 outline 美化、把讲述精髓放上屏幕。NOT FOR 未经授权改写作者内容或生成企业 PPT。

Quality

88%

Does it follow best practices?

Run evals on this skill

Adds up to 20 points to the overall score

View guide

Securityby

Passed

No findings from the security scan

Quality

Content

82%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 exceptionally concise, decision-dense overview with concrete tooling, explicit validation and honest-reporting requirements, and a clean reference structure. The main gaps are the missing BuildDeck.ts invocation and the fact that none of the referenced files were provided for verification.

Suggestions

Include the exact command for running Tools/BuildDeck.ts (or point to where its CLI is documented in RenderingSpec.md) so the build step is directly executable from the skill body.

Surface or link a minimal 5-step build→validate→deliver sequence inline (read design refs → build → static validation → browser probe → deliver), with an explicit fix-and-revalidate loop, since the detailed workflow lives in Workflows/Generate.md.

Ensure the referenced bundle files (Workflows/Generate.md, DesignSystem.md, CompositionReference.md, RenderingSpec.md, ChartSpec.md, Tools/BuildDeck.ts, Tools/ProbeDeck.js, SloganTemplate.html) ship with the skill — none were present in the provided material, leaving the reference structure unverifiable.

Dimension	Reasoning	Score
Conciseness	Lean and efficient throughout: the Gotchas ("两行可以是一个判断", "不要把中文逐字塞进固定字符格") and design rules ("不在产物末尾补 CSS，不另写渲染分支") state operational decisions with zero padding and no explanation of concepts Claude already knows. Nothing in the body could be trimmed, matching anchor 5 rather than the 'minor instances to trim' of anchor 4.	5 / 5
Actionability	Concrete guidance names the exact tool chain ("使用 Tools/BuildDeck.ts 把数据装入 SloganTemplate.html"), output path ("~/Downloads/{title}.html"), theme flags (-r / --theme=red), keybindings, and three worked request-to-mode examples. It falls short of anchor 5 because no actual invocation for BuildDeck.ts is given, so the build step is not copy-paste executable from the body alone.	4 / 5
Workflow Clarity	A sequence is legible (content contract → design references → build → deliver → "静态验证" → isolated browser probe via Tools/ProbeDeck.js → thumbnail review), and validation is explicit with a failure fallback ("隔离浏览器不可用时保留产物并如实报告未完成视觉复验"). It falls short of anchor 5 because the multi-step sequence itself is delegated to Workflows/Generate.md, so explicit error-recovery checkpoints are not visible in this file.	4 / 5
Progressive Disclosure	The body is a well-organized overview with clearly signaled references: a routing table for Workflows/Generate.md and a 设计与实现 section naming each file's role (DesignSystem.md, CompositionReference.md, RenderingSpec.md, ChartSpec.md, Tools/BuildDeck.ts). It falls short of anchor 5 because the referenced bundle files were not present alongside SKILL.md in the provided material, so the one-level-deep reference structure cannot be verified as real files.	4 / 5
	Total	17 / 20 Passed

Description

92%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.

The description is dense, concrete, and fully specifies what the skill does, when to use it, and what it explicitly will not do. Trigger coverage is good but omits several common synonyms for presentations, which is the only meaningful gap.

Dimension	Reasoning	Score
Specificity	Multiple concrete actions with comprehensive coverage: inputs ("Org/Markdown 提纲"), output ("单文件离线 HTML"), layout selection ("以语义角色选择巨句、对照、Unicode 关系图、证据与章节版式"), and delivery features ("支持讲稿层、原生图表与翻页笔"). The "Unix 气质" opener is stylistic flavor, but the rest of the description is concrete throughout, matching the anchor-5 example rather than the 'minor gaps' of anchor 4.	5 / 5
Completeness	Explicitly answers both questions: the "what" is faithful presentation design/typesetting producing single-file offline HTML, and the "USE WHEN" clause lists concrete trigger phrases, with an additional "NOT FOR" boundary. This matches the anchor-5 example, which pairs a clear capability statement with an explicit, trigger-anchored "when".	5 / 5
Trigger Term Quality	Good natural trigger coverage: "present、做演讲、slides、按 outline 美化、把讲述精髓放上屏幕" are phrases users would actually say. Missing common synonyms such as 幻灯片, deck, or presentation; "PPT" appears only in the exclusion clause, so coverage is a few terms short of the comprehensive anchor-5 example.	4 / 5
Distinctiveness Conflict Risk	Clear niche (faithful/editorial deck-building from outlines in a distinctive Unix aesthetic) and the "NOT FOR 未经授权改写作者内容或生成企业 PPT" clause fences off the most likely conflict, a generic corporate-PPT skill. Conflict risk is minimal, fitting anchor 5 rather than the 'minor overlap' of anchor 4.	5 / 5
	Total	19 / 20 Passed

Validation

87%

Warnings & errors only

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

Validation — 14 / 16 Passed

Validation for skill structure

Criteria	Description	Result
frontmatter_unknown_keys	Unknown frontmatter key(s) found; consider removing or moving to metadata	Warning
relative_links	Relative link issues: 5 missing	Warning

	Total	14 / 16 Passed

Repository: lijigang/ljg-skills
Path: skills/ljg-present/SKILL.md
Commit: 7449831

Reviewed: 2 days ago

Table of Contents

Discovery Implementation Validation

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.