Analyzes daily AI conversation logs to detect cognitive blind spots and recommends a targeted WeRead chapter.
npx skills add https://github.com/lijigang/ljg-skills --skill ljg-blind
Please help me install this Agent Skill. Skill: ljg-blind Source: https://github.com/lijigang/ljg-skills/tree/main/skills/ljg-blind Command: npx skills add https://github.com/lijigang/ljg-skills --skill ljg-blind
It reveals structural thinking habits that make certain truths systematically invisible to you—patterns you cannot see yourself—by scanning your own AI conversation history and pairing the finding with a precise book chapter to correct it.
Ask the AI to scan yesterday's conversations for my thinking blind spots and recommend a WeRead chapter to fix the top one.
High-intent search queries and answers about installing and using this skill.
Run the scan with an optional date argument; it reads Codex session JSONL files for that day, filters your actual user messages, and checks them against five blind spot signal types with quoted evidence. The output is an org-mode note naming one blind spot with supporting quotes.
It extracts core keywords from the detected blind spot, searches the WeRead API for books rated 7.5 or higher, picks the most relevant chapter from the table of contents, estimates reading time from word count, and generates an encoded web reader URL.
Yes, pass a date in YYYY-MM-DD format as an argument to scan that day's conversations instead of the default yesterday. The tool checks both the target and previous day session directories to handle timezone boundaries.
If the day's genuine user input is under 200 characters or missing entirely, the output explicitly states the data was too sparse rather than fabricating a blind spot. The skill refuses to invent findings without evidence.
Raw bookId and chapterUid values cannot be used directly in web reader URLs. They must be encoded with the bundled WeReadWebUrl.ts Bun script, which produces the /web/reader/{encodedBookId}k{encodedChapterUid} format the site accepts.
hwj123hwjIntegrates the WereAD API to browse bookshelves, search books, and view reading statistics and notes.
davekilleenOrganizes digital clutter by routing files and extracting tasks based on your goals.
haryelramalhoVerify code review feedback against the codebase before implementing changes.
studyplusCalculate Pythagorean numerology life numbers from a name.
