ljg-book

Book reader for non-specialists. Reconstructs a whole book on two independent axes: a source-blind reader can retell what the book literally follows, how its indispensable lines relate, and where it ends; the same reader can enter a concrete path in which evidence, events, or consequences revise understanding. Never let a portable thesis replace the book itself. USE WHEN given a book title, PDF, EPUB, excerpt, or asked 拆书, 分析这本书, 这本书在讲什么, 压缩一本书, or book. Defaults to a saved Org note. NOT FOR chapter summaries, papers, formal reviews, or single-concept deep dives.

Quality

88%

Does it follow best practices?

Run evals on this skill

Adds up to 20 points to the overall score

View guide

Securityby

Low

Low-risk findings worth noting

Quality

Content

77%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 well-structured with clear workflows, validation feedback loops, and real supporting files, and it largely assumes Claude's competence. Its main weakness is a missing referenced file (ReadingGuide.md) that the routing table marks as required reading, which leaves the primary navigation path incomplete.

Suggestions

Add the missing `ReadingGuide.md` (or correct the path) — the Workflow Routing table marks it as required reading for three of three input types, so its absence blocks the skill's main entry path.

Tighten the abstract process lists (Gotchas, 认识更新合同, 写作声音) by merging overlapping items; several guidance points restate the same 'let the reader feel the gap first' idea, adding tokens without new coverage.

Move the detailed delivery-gate checklists (整书四问 / 认识九问) into a separate references file and keep a short inline summary, so the SKILL.md overview stays a lean entry point.

Dimension	Reasoning	Score
Conciseness	The body is dense and largely assumes Claude's competence, with executable commands and minimal padding, but several sections (e.g. the long Gotchas and 认识更新合同 lists) restate abstract process guidance that could be tightened. It sits noticeably above the midpoint but short of lean perfection.	4 / 5
Actionability	It provides concrete, executable guidance — the `bun …/scripts/validate_note.ts … --coverage …` command, the `date +%Y%m%dT%H%M%S` timestamp recipe, the filename template, and worked prose examples — though much of the core instruction is conceptual writing-voice guidance rather than copy-paste code.	4 / 5
Workflow Clarity	A routing table sequences inputs to required reads and outputs, and the Completion/Delivery-gate sections give explicit validation checkpoints with feedback loops (run validator → read back warnings → re-validate via real Emacs `org-lint`), matching the anchor for clear sequences with error-recovery loops.	5 / 5
Progressive Disclosure	Real bundle files are appropriately referenced (`references/template.org`, `references/coverage-map.md`, `scripts/validate_note.ts` all exist), but `ReadingGuide.md` is referenced three times in the routing table as a required read and is not present in the bundle, breaking navigation for the skill's primary entry path.	3 / 5
	Total	16 / 20 Passed

Description

100%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 concise yet fully specifies what the skill does, when to use it (with multilingual trigger terms), and where it does not apply. It uses third person, avoids fluff, and is clearly distinguishable from neighboring skills.

Dimension	Reasoning	Score
Specificity	The description lists multiple concrete actions across two independent axes — retelling 'what the book literally follows, how its indispensable lines relate, and where it ends' and entering 'a concrete path in which evidence, events, or consequences revise understanding' — giving comprehensive coverage rather than a single generic verb.	5 / 5
Completeness	It explicitly answers both what ('Reconstructs a whole book on two independent axes…') and when ('USE WHEN given a book title, PDF, EPUB, excerpt, or asked…'), with concrete trigger phrases and an explicit exclusion clause.	5 / 5
Trigger Term Quality	It gives comprehensive natural triggers including file types ('PDF, EPUB, excerpt'), the bare term 'book', and several Chinese phrases users would actually say ('拆书, 分析这本书, 这本书在讲什么, 压缩一本书'), covering synonyms and extensions.	5 / 5
Distinctiveness Conflict Risk	The narrow niche ('Book reader for non-specialists' reconstructing a whole book) plus the explicit 'NOT FOR chapter summaries, papers, formal reviews, or single-concept deep dives' boundary minimizes overlap with adjacent skills.	5 / 5
	Total	20 / 20 Passed

Validation

93%

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 — 15 / 16 Passed

Validation for skill structure

Criteria	Description	Result
frontmatter_unknown_keys	Unknown frontmatter key(s) found; consider removing or moving to metadata	Warning

	Total	15 / 16 Passed

Repository: lijigang/ljg-skills
Path: skills/ljg-book/SKILL.md
Commit: 44f4105

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.