ljg-blind

盲区扫描——读昨天你与 AI 的全部对话，照出暴露的思维盲区（不是不懂的知识，是让某类真相一直看不见的思维习惯），再从微信读书挑一本书的一章精准补上，落成一篇完整分析笔记。Use when user says '扫盲区', '盲区', '照盲区', '看看我的思维盲区', '我昨天想漏了什么', 'blind spot', 'ljg-blind', '/ljg-blind', or wants yesterday's AI conversations analyzed for cognitive blind spots plus a targeted weread chapter to fill them. 可选参数：传一个日期(YYYY-MM-DD)扫那天而非昨天。NOT FOR 纵向深钻一个观点（用 ljg-think）、一个领域降秩（用 ljg-rank）、陪读一篇文本（用 ljg-read）、找一个领域的约束（用 ljg-constraint）。

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

77%Weight 40%Scale 1-3

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

A highly actionable, clearly sequenced workflow skill with strong fallbacks and self-checks. It loses points on conciseness (some conceptual and literary-style prose could be trimmed) and on progressive disclosure (everything lives in one file with no bundle references to offload detail into).

Suggestions

Tighten the non-actionable prose: condense the '盲区是什么' definitional section and the '语气规则' literary coaching (the 汪曾祺/翻译腔 tests) into terse bullets, keeping the five blind-spot signals and the self-check as the core.

Move reference-heavy material into bundle files and link from the body — e.g. references/org-syntax.md for the org-mode rules and references/weread-api.md for the curl/jq details — to enable one-level-deep progressive disclosure and shrink the main SKILL.md.

Hoist the time-sensitive skill_version value ('1.0.3') out of the inline curl examples into a single declared variable or a 'deprecated/old patterns' note so future version bumps touch one place.

Dimension	Reasoning	Score
Conciseness	Mostly efficient and dense with actionable material, but the '盲区是什么' definitional section and the literary style coaching ('一个没读过翻译小说的中国人会这么说吗？汪曾祺会这么写吗？') include explanation that could be tightened without losing the working core.	2 / 3
Actionability	Fully executable guidance throughout — concrete bash/jq filters for conversation extraction, real curl payloads for the weread API, and a complete org-mode note template with specific file paths, all copy-paste ready.	3 / 3
Workflow Clarity	Clear five-step sequence (定日期 → 捞对话 → 读盲区 → 选章 → 写笔记) with explicit error-recovery fallbacks ('数据稀薄兜底', weread-empty 兜底) and a 自检 checklist; operations are read-only/non-destructive so no missing-validation cap applies.	3 / 3
Progressive Disclosure	A monolithic ~187-line single file with no references/scripts/assets bundle; it is well-sectioned, but separable material (org-mode syntax rules, tone rules, weread API specifics) is inline and the under-50-lines simple-skill exception does not apply.	2 / 3
	Total	10 / 12 Passed

Description

100%Weight 40%Scale 1-3

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 actions, natural trigger terms, explicit what-and-when guidance, and clear disambiguation from sibling skills. Its only weakness is verbosity — the parenthetical definitional aside and the 'NOT FOR' enumeration make it longer than necessary, though verbosity is not a scored dimension here.

Dimension	Reasoning	Score
Specificity	Lists multiple concrete actions — '读昨天你与 AI 的全部对话，照出暴露的思维盲区...再从微信读书挑一本书的一章精准补上，落成一篇完整分析笔记' — in imperative/third-person voice with no first/second-person phrasing to penalize.	3 / 3
Completeness	Explicitly answers both 'what' (read conversations → surface blind spot → pick a weread chapter → write analysis note) and 'when' via a 'Use when user says...' clause, so it is not capped at 2.	3 / 3
Trigger Term Quality	Covers natural phrases a user would actually say — '扫盲区', '盲区', '照盲区', '看看我的思维盲区', '我昨天想漏了什么', 'blind spot', 'ljg-blind', '/ljg-blind' — plus a plain-language trigger about analyzing yesterday's conversations.	3 / 3
Distinctiveness Conflict Risk	Clear niche (yesterday's-conversation blind-spot scan + weread chapter) with an explicit 'NOT FOR' list disambiguating ljg-think, ljg-rank, ljg-read, and ljg-constraint, making wrong-skill triggering unlikely.	3 / 3
	Total	12 / 12 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: 1 missing	Warning

	Total	14 / 16 Passed

Repository: lijigang/ljg-skills
Path: skills/ljg-blind/SKILL.md
Commit: 0e78094

Reviewed: 9 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.