CtrlK
BlogDocsLog inGet started
Tessl Logo

ljg-read

用《学习观》的「下上结构」讲透用户提供的文本:从具体情境接通概念、判别依据与有条件的关系,让读者理解原意并能迁移判断;默认直接在对话中展示。USE WHEN 用户说「伴读」「陪我读」「读这篇」「讲透这段」「read with me」或提供文本希望深入浅出地讲解。NOT FOR 单纯翻译、整本书拆解或只画概念地图。

65

Quality

82%

Does it follow best practices?

Run evals on this skill

Adds up to 20 points to the overall score

View guide
SecuritybySnyk

Passed

No findings from the security scan

SKILL.md
Quality
Evals
Security

Quality

Content

67%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.

A well-structured, genuinely actionable instruction skill with a clear workflow order and an explicit final verification checklist. Its main weakness is redundancy: the same no-proof-of-learning caveat is repeated across four sections, which inflates token cost without adding guidance.

Suggestions

Consolidate the 「复述/单次正确回答不能证明读者已学会」 caveat into a single place (e.g., 交付检查 or Gotchas) and reference it elsewhere, removing the repetitions in 「材料和验证」 and 讲解约定.

Add one short worked example of a finished 讲解 output (a few sentences showing 情境 → 概念 → 回到情境) so the deliverable format is demonstrable, not just described.

Consider moving the 《学习观》 theory paraphrase into a reference file and keeping only the operational rules in SKILL.md, which would both shorten the body and improve progressive disclosure.

DimensionReasoningScore

Conciseness

The prose is dense rather than padded, and the 下上结构 section is legitimate domain knowledge (a specific book's framework) rather than concepts Claude already knows. However, the caveat that restated examples or one correct answer cannot prove learning (「复述已讲例子不足以证明会迁移」「一次正确回答…均不足以证明…读者已经掌握」) recurs across four sections (材料与验证, 讲解约定, Gotchas, 交付检查), which is tightening material beyond the minor level of a 4.

3 / 5

Actionability

For an instruction-only skill, the guidance is concrete: input handling with an explicit fallback (「拿不到正文时说明缺口,不根据标题或摘要补写全文」), a negative example of bad lower-layer material (「好人也会死」), output-format rules (tables only for parallel comparisons, symbols explained first), and three input→behavior Examples. It misses a sample of a finished 讲解 output, which keeps it from fully copy-paste-ready 5.

4 / 5

Workflow Clarity

A clear sequence runs from Workflow Routing → 原文与范围 (acquire text, fallback) → 把下层与上层接通 (一句点明 → 情境进入 → 概念关系 → 回到情境) → 对话成品 (format) → 可选交互 → 交付检查 (an explicit end-of-task verification checklist). Not a destructive or batch operation, so no cap applies; some intermediate checkpoints are implicit rather than explicit, keeping it below 5.

4 / 5

Progressive Disclosure

No bundle files exist and the body references none, so there are no broken or nested references. The ~75-line body is well-sectioned with a routing table, clear headers, and appropriately scoped sections; it slightly exceeds the under-50-line simple-skill case and the theory section could plausibly live in a reference file, which is the minor gap below 5.

4 / 5

Total

15

/

20

Passed

Description

87%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: it states concrete capabilities, gives explicit USE WHEN triggers in natural user language (including an English variant), and demarcates boundaries with a NOT FOR clause. Minor room to grow on trigger synonym coverage and surface-level capability enumeration.

DimensionReasoningScore

Specificity

The description names a concrete domain (讲透用户文本 via the 下上结构 methodology) and several specific actions — 「从具体情境接通概念、判别依据与有条件的关系」「让读者理解原意并能迁移判断」「默认直接在对话中展示」. It stops short of 5 because coverage is not fully comprehensive (e.g., file/URL input handling is only in the body), but it well exceeds the 1–2-action level of a 3.

4 / 5

Completeness

It explicitly answers both what (用「下上结构」讲透文本、接通概念与判别依据、让读者能迁移判断、默认在对话中展示) and when (「USE WHEN 用户说『伴读』『陪我读』『读这篇』『讲透这段』『read with me』或提供文本希望深入浅出地讲解」), with concrete trigger phrases and an additional NOT FOR boundary — a textbook match for the score-5 anchor.

5 / 5

Trigger Term Quality

Triggers are natural phrases a user would actually say: 「伴读」「陪我读」「读这篇」「讲透这段」「read with me」, plus the fallback condition 提供文本希望深入浅出地讲解. A few common variants (e.g., 「帮我理解这段」「解读一下」) are missing, keeping it below the comprehensive 5.

4 / 5

Distinctiveness Conflict Risk

A clear niche (guided reading using a specific book's methodology) with distinct colloquial triggers and an explicit NOT FOR clause excluding 单纯翻译、整本书拆解、只画概念地图, minimizing overlap with translation, book-summary, or diagramming skills.

5 / 5

Total

18

/

20

Passed

Validation

100%

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

Validation — 16 / 16 Passed

Validation for skill structure

No warnings or errors.

Repository
lijigang/ljg-skills
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.