daftai-chinese-copywritinglisted
Install: claude install-skill daftAI2026/daftAI-skills
# Chinese Copywriting
Checks and fixes Chinese copywriting with `autocorrect`, using the upstream guideline as the rule source.
## Script Directory
**Important**: All scripts are located in the `scripts/` subdirectory of this skill.
**Agent Execution Instructions**:
1. Determine this SKILL.md file's directory path as `SKILL_DIR`
2. Script path = `${SKILL_DIR}/scripts/<script-name>.ts`
3. Execute via `bun ${SKILL_DIR}/scripts/<script-name>.ts`
4. Replace all `${SKILL_DIR}` in this document with actual values
**Script Reference**:
| Script | Purpose |
|--------|---------|
| `scripts/main.ts` | CLI entry for review/stable/quick workflows |
| `scripts/autocorrect.ts` | Detect, install, and run `autocorrect` |
| `scripts/shared.ts` | Preferences, input handling, and syntax protection (code/math/links) |
| `scripts/chunk.ts` | Markdown-aware chunking for long texts |
All scripts are executed via `bun`. `scripts/chunk.ts` depends on `markdown-it` which is installed under `scripts/node_modules/`. Before AI post-processing, `protectSyntax` replaces fenced code blocks, inline code, block/inline math (`$$`/`$`), and link/image URLs with placeholders; `restoreSyntax` reverses this after processing.
## Defaults
| Setting | Default | EXTEND.md key | Description |
|---------|---------|---------------|-------------|
| Mode | `stable` | `default_mode` | Workflow mode (`stable` or `quick`; `review` is inferred from user intent) |
| Chunk threshold | `4000` | `chunk_threshold` | Word co