cnblogs-publisher

name	cnblogs-publisher
description	Publish, update, inspect, and organize local Markdown articles on cnblogs.com through the MetaWeblog API and an optional Playwright collection workflow. Use when the user explicitly asks to publish or update an article on CNBlogs, list or inspect existing CNBlogs posts, or add posts to a CNBlogs collection.

CNBlogs Publisher

Publish local Markdown files through the CNBlogs MetaWeblog API. Collection membership is managed separately through the CNBlogs web editor because the MetaWeblog API does not expose that setting.

Safety

• Publishing and updating are external, user-visible mutations. Only run those commands after the user has explicitly requested the corresponding action.
• Confirm the target file, title, categories, tags, and update post ID before publishing or replacing an existing post.
• Never put credentials or browser session data in source files, command arguments, logs, or chat output.
• Keep .env and .browser_profile/ untracked. The repository .gitignore already excludes both names at any directory depth.
• publish.py publishes immediately. Use it only when a public post is intended.

Configuration

The scripts preserve already exported environment variables and optionally load the nearest .env found by walking upward from the current directory and the script directory.

Required for MetaWeblog operations:

Variable	Purpose
CNBLOGS_ENDPOINT	MetaWeblog XML-RPC endpoint
CNBLOGS_USERNAME	CNBlogs login name
CNBLOGS_PASSWORD	MetaWeblog access token or password
CNBLOGS_BLOGID	Blog ID passed to MetaWeblog

Optional publishing defaults:

Variable	Purpose	Default
CNBLOGS_BLOG_URL	Public blog base URL used to print post links	https://www.cnblogs.com/<CNBLOGS_BLOGID>
CNBLOGS_CATEGORIES	Comma-separated post categories	[Markdown]
CNBLOGS_TAGS	Comma-separated post tags	empty

Optional collection defaults:

Variable	Purpose
CNBLOGS_COLLECTION_ID	Collection checkbox ID
CNBLOGS_TITLE_PATTERN	Regex used to select posts for bulk collection updates
CNBLOGS_PROFILE_DIR	Playwright persistent profile directory

Example .env with placeholders:

CNBLOGS_ENDPOINT=https://rpc.cnblogs.com/metaweblog/YOUR_BLOG_ID
CNBLOGS_USERNAME=YOUR_USERNAME
CNBLOGS_PASSWORD=YOUR_ACCESS_TOKEN
CNBLOGS_BLOGID=YOUR_BLOG_ID
CNBLOGS_BLOG_URL=https://www.cnblogs.com/YOUR_BLOG_ID
CNBLOGS_CATEGORIES=[Markdown],Python
CNBLOGS_TAGS=Python,Algorithms
CNBLOGS_COLLECTION_ID=YOUR_COLLECTION_ID
CNBLOGS_TITLE_PATTERN=^LeetCode-

Markdown Contract

publish.py treats the first line as the title and publishes the remaining lines as the body. The first line should therefore be a Markdown H1:

# Article title

Article body...

The script removes the leading # from the title and does not render Jekyll front matter. Convert Jekyll articles to this simple format before publishing.

Publish And Update

Script path:

.claude/skills/cnblogs-publisher/publish.py

Publish with environment defaults:

python3 .claude/skills/cnblogs-publisher/publish.py path/to/article.md

Override categories and tags for one post:

python3 .claude/skills/cnblogs-publisher/publish.py path/to/article.md \
  --categories "[Markdown],Python" \
  --tags "Python,Dynamic Programming"

Update an existing post only after confirming its post ID:

python3 .claude/skills/cnblogs-publisher/publish.py path/to/article.md \
  --update POST_ID

Read-only inspection commands:

python3 .claude/skills/cnblogs-publisher/publish.py --list
python3 .claude/skills/cnblogs-publisher/publish.py --list 20
python3 .claude/skills/cnblogs-publisher/publish.py --get POST_ID

After a successful publish or update, report the returned post ID and the URL printed by the script. Do not infer an account-specific URL in the skill.

Collection Membership

Script path:

.claude/skills/cnblogs-publisher/add_to_collection.py

Install the optional dependency once:

python3 -m pip install playwright
python3 -m playwright install chromium

Create a local authenticated browser profile:

python3 .claude/skills/cnblogs-publisher/add_to_collection.py --login

Add explicit posts to a collection. The collection can come from CNBLOGS_COLLECTION_ID or --collection:

python3 .claude/skills/cnblogs-publisher/add_to_collection.py \
  --postids POST_ID_A POST_ID_B \
  --collection COLLECTION_ID

For a bulk operation, a title regex is required. Preview the selection before removing --dry-run:

python3 .claude/skills/cnblogs-publisher/add_to_collection.py \
  --title-pattern '^LeetCode-' \
  --limit 100 \
  --collection COLLECTION_ID \
  --dry-run

Then run the same command without --dry-run only after the user confirms the selected posts. --headless is available after the login profile has been created.

Operational Notes

• editPost may remove collection membership. Re-run the collection command for that post after an update when collection membership must be preserved.
• The Playwright workflow uses public CNBlogs editor URLs and locates the collection checkbox by its configured numeric ID.
• Website markup can change. Use --dry-run and a small explicit post list when validating the automation after a CNBlogs UI change.
• Platform rate limits are external state. Stop on a limit response and report completed and remaining posts instead of retrying publication indefinitely.