wewrite-learn

Learns writing preferences from user edits, exemplar articles, and WeChat layout themes.

3.2k|513|Updated Mar 26, 2026
One-click install
npx skills add https://github.com/oaker-io/wewrite --skill wewrite-learn
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: wewrite-learn
Source: https://github.com/oaker-io/wewrite/tree/main/skills/wewrite-learn
Command: npx skills add https://github.com/oaker-io/wewrite --skill wewrite-learn

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill includes references (resource) components.

What problem does it solve?

Writing assistants produce drafts that repeatedly need the same manual corrections, and there is no systematic way to capture a user's style preferences, build a reusable exemplar library, or replicate the visual layout of existing WeChat official account articles.

Core Features & Use Cases

  • Edit Flywheel (Playbook): Compares your original draft against your manually edited final version, records recurring modification patterns as scoped rules, and promotes only repeated or explicitly confirmed rules into hard constraints for future drafts.
  • Exemplar Library: Imports reference articles from URLs or local files to calibrate structure and rhythm in future drafts, with ownership and authenticity metadata preventing reuse of personal experiences or third-party content.
  • Theme Learning: Extracts a new layout theme from any WeChat article URL so the formatting engine can render future articles in that style.
  • Use Case: After editing a generated article, say "learn my edits" — the skill diffs draft and final versions, stores patterns like shorter paragraphs for tutorials, and applies them the next time you write a similar piece.

Quick Start

Tell the agent "learn my edits" after revising a draft, or provide an article URL and ask it to import the piece as an exemplar or learn its layout theme.

Frequently Asked Questions about wewrite-learn

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I teach an AI writing assistant my editing preferences?

Run the learn-edits flow by providing the original draft and your manually edited final version. The skill records each meaningful change as a scoped pattern, and only repeated or explicitly confirmed patterns become hard constraints in future drafts.

How to import exemplar articles for style learning?

Provide a URL or local file path and ask to import it as an exemplar. The skill fetches the article, stores it with category and ownership metadata, and injects it into future draft prompts to calibrate structure and rhythm only.

Can it learn the layout theme from an existing WeChat article?

Yes. Provide the article URL with a learn-theme request, and the CLI extracts a reusable theme. After extraction, set the theme field in your style configuration to apply it to future articles.

Does a single edit become a permanent writing rule?

No. A single modification is always a soft reference. A rule becomes a hard constraint only when it repeats at least twice in the same scope with sufficient confidence, or when you explicitly confirm it as a long-term preference.

Can imported exemplars reuse personal stories from old articles?

No. All exemplars are restricted to style and structure usage, and personal materials are marked non-reusable. Even your own past articles require you to re-provide or confirm personal experiences for each new task.