WeRead Plus Companion

Problem Background

WeRead Agent API exposes data such as search, bookshelf, notes, reviews, and reading statistics, but raw endpoint assembly can turn simple tasks into fragile logic. Pagination, scoring, exports, and author-field extraction all need stable rules, while mixing personal notes with public comments can easily cross privacy boundaries. weread-plus is not a replacement for official weread-skills; it is a higher-level workflow layer built on top of it.

How It Works

The skill packages common tasks into scripts and reference docs:

• Next-read recommendations: use scripts/weread_recommend.py to generate candidates, then explain fit, mismatch, safe/expand/challenge mode, and shelf status.
• Pre-read analysis: combine book info, public reviews, popular highlights, and similar books to decide whether a book deserves time.
• Reviews and thoughts: use scripts/weread_reviews.py to fetch public reviews, review details, and popular-highlight thoughts, showing only API-returned author fields.
• Note exports: use scripts/weread_notes_export.py to output highlights and personal thoughts as Markdown or JSON.
• Reading reports: use scripts/weread_report.py for weekly, monthly, annual, overall, and bookshelf planning reports.
• API checks: use scripts/weread_call.py for low-level endpoint checks and scripts/weread_verify.py after install or upgrades.

It reads official reference files such as search.md, book.md, shelf.md, notes.md, and review.md to avoid redefining API behavior.

Boundaries

It fits workflows that turn WeRead data into reading decisions, reports, or exported notes, not tasks that bypass official access, infer private identity, or scrape non-public content. Output should separate facts from interpretation; recommendation scores are ranking aids, not objective judgments. When showing personal notes, review authors, or exported content, quote only necessary fields and do not infer identity from userVid, avatar, nickname, or writing style.