WeChat Full Publish Pipeline

Fully Automatic Mode: From topic selection to draft box, no interruptions during the process. After receiving the instruction, execute the following steps in order, no waiting for user confirmation throughout the whole process.

⚡ Progress Reporting Specification (Mandatory Execution)

After completing each Step and before invoking the tool for the next Step, output a line of progress report text first. The text and the tool invocation for the next step are in the same response, in the following order:

[Text: ✅ Step N completed: ...]
[Tool invocation: First tool for Step N+1]

This is the only way for users to see real-time progress. Do not wait for user replies, do not "end this round and wait for feedback", send the text and tool invocation together.

Report Format:

✅ Step 0 completed: Selected topic 「{title}」 (HN {N} points)
✅ Step 1 completed: Article finished, {word count} words
✅ Step 2 completed: All 5 illustrations generated
✅ Step 3 completed: Cover image generated
✅ Step 4 completed: Typesetting finished
✅ Step 5 completed: Published to draft box, media_id: {id}

Pipeline Steps

Step 0: Topic Selection        (web-search + judgment)
Step 1: Research & Write       (url-to-markdown + translate + write)
Step 2: Article Illustrations  (baoyu-article-illustrator)
Step 3: Cover Image            (baoyu-imagine)
Step 4: Markdown → HTML        (baoyu-markdown-to-html)
Step 5: Publish to Draft Box   (baoyu-post-to-wechat API)

Send a progress message to the user immediately after each step is completed (refer to the progress reporting specification above).

Step 0: Topic Selection (Autonomous Selection)

Fully autonomous, no inquiry to the user.

0.1 Hotspot Collection

Search the following sources at the same time, fetch content from the last 24-48 hours:

- Hacker News Hot List Top 20（https://news.ycombinator.com/best）
- GitHub Trending（https://github.com/trending）
- If necessary: Twitter/X AI/tech topics

0.2 Filtering Criteria

Prioritize topics that meet the following conditions:

Criteria	Description
Relevance	AI, development tools, technology trends, security incidents, engineering practices
Timeliness	Event occurred within 48 hours, with first-hand information available
Depth Potential	Has technical details to explore, not pure news retelling
Audience Match	Suitable for technology-focused Official Account readers (engineers, AI practitioners)
Not Previously Written	Check `/agent/articles/` to avoid duplicate topics

0.3 Autonomous Decision Making

Select the highest scored topic from the candidates, record the reason for topic selection, and directly proceed to Step 1. Do not display the candidate list, do not wait for feedback.

📢 Output
✅ Step 0 completed: Selected topic 「{title}」 (HN {N} points, source: {URL})
, then immediately start the tool invocation for Step 1 (crawl the original text) in the same response.

Step 1: Research & Write

1.1 Obtain Original Materials

Use

baoyu-url-to-markdown

to crawl the main source page:

bash

SKILL_DIR="/agent/wechat-agent-skills/skills/baoyu-url-to-markdown"
npx -y bun "$SKILL_DIR/scripts/main.ts" --url "{source_url}" --output /tmp/source.md

If there are multiple key sources (original text + discussion threads), crawl all of them.

1.2 Translation If Needed

If the original text is in English, use

baoyu-translate

to accurately translate the core content as reference (do not directly send the translated text):

bash

SKILL_DIR="/agent/wechat-agent-skills/skills/baoyu-translate"
npx -y bun "$SKILL_DIR/scripts/main.ts" /tmp/source.md --mode normal --to zh

1.3 Write Original Article

Article Style Requirements:

First-person narrative perspective ("That day...", "Let's start from...") or third-person narrative
Hook readers with a specific scenario/event at the beginning, do not start with "background"
Structured: opening hook → core event/concept → technical details → impact/significance → conclusion
Have your own opinions, not just news retelling
Code examples, data, citations must be accurate, from original sources
Length: 2000-4000 words
Language: Chinese

File Naming and Storage:

/agent/articles/drafts/{slug}-{YYYY-MM-DD}.md

slug: English kebab-case, 2-4 words, derived from the article topic

Frontmatter:

yaml

---
title: 文章标题
date: YYYY-MM-DD
source: 原文URL
tags: [tag1, tag2]
digest: 文章摘要（≤15字，发布时传给微信）
---

Abstract Requirements:

No more than 15 words
Point out the core event/conclusion, not a retelling of the title
Example: "AI supply chain hijacked, thousands of apps silently compromised"

📢 Output
✅ Step 1 completed: Article finished, about {word count} words, saved to {path}
, then immediately start the tool invocation for Step 2 (create illustration prompts) in the same response.

Step 2: Article Illustrations

Skill:

baoyu-article-illustrator

2.1 Analyze Article, Plan Illustrations

Read the article, judge:

Content type → select preset
- Technology/security/AI →
```
tech-explainer
```
  (infographic + blueprint)
- Narrative + technology mixed →
```
tech-explainer
```
  , use
```
timeline
```
  type for key events
- Methodology/framework →
```
system-design
```
Density:
```
per-section
```
(at least 1 image per main chapter)
Create
```
{article-dir}/imgs/outline.md
```

Good Illustration Positions:

Before the start of the article (timeline or overall overview)
After explanation of key technical mechanisms (flowchart/framework diagram)
Data/comparison sections (infographic/comparison)
Summary illustration before conclusion

2.2 Create All Prompt Files (BLOCKING, must all be created before generation)

Store to

{article-dir}/imgs/prompts/NN-{type}-{slug}.md

Prompt Quality Requirements:

```
Layout
```
: Describe the overall composition
```
ZONES
```
: Each zone has specific content, no ambiguity
```
LABELS
```
: Use actual numbers/terms/quotes from the article
```
COLORS
```
: Hex color value + semantic description
```
STYLE
```
: Use blueprint/dark/digital style for technology/security content
```
ASPECT
```
: 16:9

Content Review Security Rules (avoid triggering OpenAI content policies):

❌ "fork bomb", "malware install", "exploit", "hack", "kill process"
✅ Change to describe behavior/result: "process multiplication", "malicious payload", "vulnerability abuse mechanism"
If an image is rejected with
```
moderation_blocked
```
, immediately rewrite the prompt to describe behavior instead of technical terms, retry individually

2.3 Batch Generation

Create

{article-dir}/imgs/batch.json

json

{
  "jobs": 3,
  "tasks": [
    {
      "id": "01-slug",
      "promptFiles": ["prompts/01-type-slug.md"],
      "image": "01-type-slug.png",
      "provider": "openai",
      "model": "gpt-image-1",
      "ar": "16:9",
      "quality": "2k"
    }
  ]
}

bash

cd {article-dir}/imgs
SKILL_DIR="/agent/wechat-agent-skills/skills/baoyu-imagine"
npx -y bun "$SKILL_DIR/scripts/main.ts" --batchfile batch.json --jobs 3

2.4 Insert Into Article

Insert each image after the corresponding paragraph:

markdown

![description text](imgs/NN-type-slug.png)

📢 Output
✅ Step 2 completed: All 5 illustrations generated and inserted into the article
, then immediately start the tool invocation for Step 3 (generate cover image) in the same response.

Step 3: Cover Image

Store the cover image to

{article-dir}/imgs/cover.png

(read from this path by default by

wechat-api.ts

3.1 Create Cover Prompt

Store to

{article-dir}/imgs/prompts/cover.md

Cover Selection for Technology/AI Articles:

```
type: conceptual
```
```
palette: dark
```
```
rendering: digital
```
Mood: bold, high contrast
Color scheme: Dark navy blue/black background + electric red-orange (sense of threat) or cyan-blue (sense of technology)
No text (WeChat will display the title separately)
The main visual element must be strongly related to the article topic (do not use generic technology images)

3.2 Generate Cover

bash

cd {article-dir}/imgs
SKILL_DIR="/agent/wechat-agent-skills/skills/baoyu-imagine"
npx -y bun "$SKILL_DIR/scripts/main.ts" \
  --promptfiles "prompts/cover.md" \
  --image "cover.png" \
  --provider openai \
  --model gpt-image-1 \
  --ar 16:9 \
  --quality 2k

📢 Output
✅ Step 3 completed: Cover image generated
, then immediately start the tool invocation for Step 4 (typesetting) in the same response.

Step 4: Markdown → HTML (Typesetting)

bash

cd {article-dir}
SKILL_DIR="/agent/wechat-agent-skills/skills/baoyu-markdown-to-html"
npx -y bun "$SKILL_DIR/scripts/main.ts" {article-file} --theme grace --cite

Default configuration:

Theme:
```
grace
```
(elegant, suitable for technical content)
```
--cite
```
: Convert external links to bottom references (WeChat friendly)

Confirm that

contentImages

in the output JSON contains all illustrations before proceeding.

📢 Output
✅ Step 4 completed: Typesetting finished
, then immediately start the tool invocation for Step 5 (publish to WeChat) in the same response.

Step 5: Publish to WeChat Draft Box

bash

cd {article-dir}
SKILL_DIR="/agent/wechat-agent-skills/skills/baoyu-post-to-wechat"
npx -y bun "$SKILL_DIR/scripts/wechat-api.ts" \
  {article-file} \
  --theme grace \
  --color blue \
  --cover imgs/cover.png

Key Notes:

Pass the
```
.md
```
file, do not pass
```
.html
```
(the script handles conversion and image upload internally)
Images are automatically compressed to <1MB JPEG before upload
Return
```
media_id
```
after success

📢 Final Report: Send
✅ Step 5 completed: Published to draft box, media_id: {media_id}
, pipeline ends.

First Run Environment Check

EXTEND.md Check and Automatic Creation

bash

test -f "$HOME/.baoyu-skills/baoyu-article-illustrator/EXTEND.md" || mkdir -p "$HOME/.baoyu-skills/baoyu-article-illustrator" && cat > "$HOME/.baoyu-skills/baoyu-article-illustrator/EXTEND.md" << 'EOF'
---
version: 1
watermark:
  enabled: false
  content: ""
  position: bottom-right
  opacity: 0.7
preferred_style:
  name: null
  description: ""
default_output_dir: imgs-subdir
language: zh
custom_styles: []
---
EOF

test -f "$HOME/.baoyu-skills/baoyu-imagine/EXTEND.md" || mkdir -p "$HOME/.baoyu-skills/baoyu-imagine" && cat > "$HOME/.baoyu-skills/baoyu-imagine/EXTEND.md" << 'EOF'
---
version: 1
default_provider: openai
default_quality: 2k
default_aspect_ratio: null
default_image_size: null
default_model:
  google: null
  openai: gpt-image-1
  azure: null
  openrouter: null
  dashscope: null
  minimax: null
  replicate: null
---
EOF

test -f "$HOME/.baoyu-skills/baoyu-cover-image/EXTEND.md" || mkdir -p "$HOME/.baoyu-skills/baoyu-cover-image" && cat > "$HOME/.baoyu-skills/baoyu-cover-image/EXTEND.md" << 'EOF'
---
version: 1
watermark:
  enabled: false
  content: ""
  position: bottom-right
  opacity: 0.7
preferred_type: null
preferred_palette: null
preferred_rendering: null
preferred_text: title-only
preferred_mood: balanced
preferred_font: clean
default_aspect: 16:9
default_output_dir: imgs-subdir
quick_mode: false
language: zh
custom_palettes: []
---
EOF

test -f "$HOME/.baoyu-skills/baoyu-post-to-wechat/EXTEND.md" || mkdir -p "$HOME/.baoyu-skills/baoyu-post-to-wechat" && cat > "$HOME/.baoyu-skills/baoyu-post-to-wechat/EXTEND.md" << 'EOF'
default_theme: grace
default_color: blue
default_publish_method: api
default_author: 宝玉
need_open_comment: 1
only_fans_can_comment: 0
EOF

npm Dependency Check

bash

ls /agent/skills/baoyu-skills/node_modules/jimp 2>/dev/null || \
  (cd /agent/skills/baoyu-skills && npm install jimp @jsquash/webp)

Completion Report

WeChat Full Pipeline Complete!

Topic: [topic]
Source: [source URL]
Article: [article path]
Illustrations: X pieces
Cover: imgs/cover.png
Theme: grace + blue

✓ Published to Official Account draft box
• media_id: [media_id]

→ Draft Box: https://mp.weixin.qq.com (Content Management → Draft Box)

wechat-publish-pipeline

NPX Install

Tags

SKILL.md Content (Chinese)

WeChat Full Publish Pipeline

⚡ Progress Reporting Specification (Mandatory Execution)

Pipeline Steps

Step 0: Topic Selection (Autonomous Selection)

0.1 Hotspot Collection

0.2 Filtering Criteria

0.3 Autonomous Decision Making

Step 1: Research & Write

1.1 Obtain Original Materials

1.2 Translation If Needed

1.3 Write Original Article

Step 2: Article Illustrations

2.1 Analyze Article, Plan Illustrations

2.2 Create All Prompt Files (BLOCKING, must all be created before generation)

2.3 Batch Generation

2.4 Insert Into Article

Step 3: Cover Image

3.1 Create Cover Prompt

3.2 Generate Cover

Step 4: Markdown → HTML (Typesetting)

Step 5: Publish to WeChat Draft Box

First Run Environment Check

EXTEND.md Check and Automatic Creation

npm Dependency Check

Completion Report