mlwcli（mlwcli 搜索）

mlwcli

A unified CLI for managing feeds (Miniflux), links (Linkding), and pages (Wallabag).

Command Reference

# Authentication (interactive TUI)
mlwcli auth login        # Login to linkding, miniflux, or wallabag
mlwcli auth logout       # Logout from a service

# Linkding (Links)
mlwcli link add <url>    # Add link
mlwcli link list         # List links

# Miniflux (Feeds)
mlwcli feed add <url>    # Add feed
mlwcli feed list         # List feeds
mlwcli entry list        # List feed entries
mlwcli entry save <id>   # Save entry to third-party service

# Wallabag (Pages)
mlwcli page add <url>    # Add page
mlwcli page list         # List pages

Use --help on any command for options.

Critical Guidelines

1. Authentication requires user action:

- auth login and auth logout use an interactive TUI that requires a TTY — do not run them yourself - If a command fails due to missing credentials, tell the user to run mlwcli auth login in their terminal - Config stored at ~/.config/mlwcli/auth.toml

1. Pagination: All list commands return {total, items}.

- link list, entry list: --limit / --offset (default: limit=10, offset=0) - page list: --page / --per-page (default: page=1, per-page=10) - feed list: no pagination parameters (returns all) - Always check total vs items returned before assuming you have all results.

1. Output Filtering:

- --jq=expression — inline jq filter (automatically enables JSON output) - --json=field1,field2 — select specific fields (comma-separated, no spaces) - Without these flags, output is human-readable table format

1. Search and Filtering:

- link list: --search, --limit, --offset - entry list: --search, --status (read/unread/removed, default: unread), --starred, --feed-id, --limit, --offset - page list: --archive, --starred, --tags, --domain, --page, --per-page

1. Quote Handling:

- Values with double quotes: wrap in single quotes: --notes 'Title: "Example"' - Tags are space-separated within a quoted string: --tags "tag1 tag2"

Examples

Authentication

Authentication requires an interactive TTY — ask the user to run these commands themselves:

mlwcli auth login   # interactive TUI: select service, enter endpoint + credentials
mlwcli auth logout  # interactive TUI: only shows signed-in services

Check pagination before processing

mlwcli entry list --status=unread --jq='{total: .total, returned: (.items | length)}'

If total > returned, increase --limit or paginate:

mlwcli entry list --status=unread --limit=100
# or iterate with --offset=0, --offset=10, etc.

List unread entries

mlwcli entry list --status=unread --jq='.items[] | { id, url, title, published_at, status, feed_id: .feed.id, feed_title: .feed.title }'

List entries by feed

# Find feed ID
mlwcli feed list --jq='.items[] | { id, title, site_url }'

# Fetch entries from that feed
mlwcli entry list --feed-id=42 --limit=20 --jq='.items[] | { id, url, title, published_at }'

Find starred/read entries by date

mlwcli entry list --starred --status=read --limit=100 --json=id,url,title,changed_at,starred | jq '.items[] | select(.changed_at >= "2025-12-26")'

Note: changed_at reflects when the entry was last modified (starred/read status changed), not publication date.

Save an entry to a third-party service

mlwcli entry save 42

Saves to Miniflux's configured third-party integration (e.g. Wallabag, Pocket).

Add a feed (with optional category)

mlwcli feed add <url>
mlwcli feed add <url> --category-id=<id>   # defaults to 1 (All) if omitted

To find category IDs:

mlwcli feed list --jq='.items[] | { id, title, category_id: .category.id, category_title: .category.title }'

Add a link

mlwcli link add <url>
mlwcli link add <url> --notes='Title: "Some Title"' --tags="tag1 tag2"

Add a page

mlwcli page add <url>
mlwcli page add <url> --tags="tag1 tag2" --archive

List pages

mlwcli page list --starred --per-page=20 --jq='.items[] | { id, url, title, domain_name }'
mlwcli page list --domain=example.com
mlwcli page list --tags="tech news"