xinzhi

name	xinzhi
description	Manage notes, containers, tags, and authors on Xinzhi (xinzhi.zone) via the xinzhi CLI. Use when the user asks to save a link, create a memo, write a document, search notes, manage tags or containers, list authors, archive or move notes, or anything related to knowledge management, bookmarking, or note-taking with Xinzhi.

Xinzhi Knowledge Manager

Interact with the user's Xinzhi account via the xinzhi CLI (v0.0.9+) to manage notes, containers, tags, and authors.

Prerequisites

• xinzhi CLI installed globally: npm install -g @xinzhi-zone/cli
• Logged in: xinzhi login <token> (token from Xinzhi web Settings -> AI Agent -> CLI AccessToken)
• Node.js >= 20

Quick Reference

All commands output JSON. Parse and present results in a user-friendly format.

Containers (folders for organizing notes)

# List all containers
xinzhi container list --page-size 50

# Create a container
xinzhi container create --title "AI Inbox" --description "Agent Inbox"

# Get by name or ID
xinzhi container get --name "AI Inbox"
xinzhi container get --id "<container-id>"

# Update
xinzhi container update --id "<id>" --title "New Title" --description "New Desc"

Tags

# List tags
xinzhi tag list --keyword "AI" --page-size 50

# Create a tag (supports hierarchical names like "AI/Product")
xinzhi tag create --name "AI/Product"

# Get by name or ID
xinzhi tag get --name "AI/Product"

# Update
xinzhi tag update --id "<id>" --name "AI/Engineering"

Authors

# List authors
xinzhi author list --keyword "张" --page-size 20

Notes

Creating Notes

# Save a web link
xinzhi note create-link --url "https://example.com" \
  --container-name "AI Inbox" --tag-names "AI/Product"

# Quick memo
xinzhi note create-memo --content "Today's idea..." \
  --container-name "AI Inbox" --tag-names "AI/Product,Thinking"

# Long-form document (markdown)
xinzhi note create-document --title "Weekly Summary" \
  --content "$(cat ./weekly-summary.md)" \
  --container-name "AI Inbox" --tag-names "AI/Product"

# Highlight a text fragment in a document note
xinzhi note create-highlight --note-id "<document-note-id>" \
  --text "selected text" --prefix "text before" --suffix "text after" \
  --highlight default --highlight-color yellow

# Annotate a note (full-note thought)
xinzhi note create-annotation --note-id "<note-id>" --content "My thoughts"

# Annotate a specific quote in a document note
xinzhi note create-annotation --note-id "<note-id>" \
  --content "My annotation" \
  --text "selected text" --prefix "text before" --suffix "text after"

Querying Notes

# List notes with filters
xinzhi note list --container-name "AI Inbox" --tag-name "AI/Product" --page-size 20

# List with time range and sorting
xinzhi note list --start-time "2026-01-01" --end-time "2026-03-31" \
  --time-field publishTime --type ARTICLE --sort-by publishTime --sort desc

# Full-text search (own notes)
xinzhi note search --keyword "Agent" --type ARTICLE --page-size 20

# Search subscription/newsletter articles (oneself=false in results)
xinzhi note search --keyword "参考答案" --type SUBSCRIBE --page-size 50

# Aggregate search (returns articleSearchResult, memoSearchResult,
# highlightSearchResult, and subscribeSearchResult in one call)
xinzhi note aggregate-search --keyword "Agent" --page-size 20

# Get single note by ID
xinzhi note get --id "<note-id>"

Managing Note Tags

# Add a tag
xinzhi note add-tag --id "<note-id>" --tag-name "AI/Product"

# Remove a tag
xinzhi note remove-tag --id "<note-id>" --tag-name "AI/Product"

# Replace all tags
xinzhi note set-tags --id "<note-id>" --tag-names "AI/Product,Thinking"

# Clear all tags
xinzhi note set-tags --id "<note-id>" --clear-tags

Organizing Notes

# Move to another container
xinzhi note move --id "<note-id>" --to-container-name "Inbox"

# Archive / Unarchive
xinzhi note archive --id "<note-id>"
xinzhi note unarchive --id "<note-id>"

Workflow Patterns

Save and Tag a Link

1. Create the link note: xinzhi note create-link --url "<url>" --container-name "<container>"
2. Parse the returned JSON for the note id
3. If additional tags needed: xinzhi note add-tag --id "<id>" --tag-name "<tag>"

Batch Tag Management

1. Search for notes: xinzhi note search --keyword "<keyword>"
2. For each matching note, apply tags: xinzhi note set-tags --id "<id>" --tag-names "<tags>"

Find Notes by Author

1. Find the author: xinzhi author list --keyword "<name>"
2. Extract authorId from results
3. List their notes: xinzhi note list --author-id "<authorId>"

Search Subscription Articles

To find articles from a specific newsletter/subscription source:

1. Search by source name with SUBSCRIBE type: xinzhi note search --keyword "<source-name>" --type SUBSCRIBE --page-size 50
2. Or use aggregate search: xinzhi note aggregate-search --keyword "<source-name>" and check subscribeSearchResult
3. Distinguish ownership: oneself=true means user-created notes, oneself=false means subscription-sourced articles

Create a Document from Content

When the user provides long-form content or asks to save a document:

1. Format content as markdown
2. Use $'...' quoting or "$(cat file.md)" for multi-line content
3. Create: xinzhi note create-document --title "<title>" --content "$content" --container-name "<container>"

Important Notes

• All output is JSON; parse it before presenting to the user
• Container and tag options support both --*-id and --*-name (mutually exclusive)
• Tag names support hierarchy via / separator (e.g., "AI/Product")
• Note types for search --type: ARTICLE (own links/documents), MEMO, HIGHLIGHT, SUBSCRIBE (newsletter articles)
• oneself field in results: true = user-created, false = from subscriptions
• aggregate-search returns separate result sets: articleSearchResult, memoSearchResult, highlightSearchResult, subscribeSearchResult
• Time fields for filtering: createTime, editTime, publishTime
• Sort fields: title, createTime, editTime, publishTime, readProgress
• For markdown content in shell, use $'...' syntax or read from a file to handle newlines correctly
• Highlight and annotation commands require an existing note ID
• Highlight text must be an exact fragment from the original document