slax-reader — slax-lab/slax-reader-cli Agent Skill

Mode d’emploi

Copiez la commande et examinez le paquet avant de l’installer dans votre environnement agent.
Consultez le document original pour les outils, identifiants et environnements requis.
Décrivez les entrées et le livrable souhaité ; comparez la sortie aux sources.

npx skills add https://github.com/slax-lab/slax-reader-cli --skill slax-reader

Avant de commencer

L’installation n’inclut ni abonnement au modèle, ni quota API, ni accès aux fichiers.
Étoiles et installations reflètent l’activité, pas un test indépendant de qualité ou de sécurité.

Document Skill original

Rendu depuis le dépôt source en conservant titres, exemples, code, tableaux, liens et images.

Slax Reader CLI

Use reader-cli to save URLs into the user's Slax Reader library, list saved bookmarks, and get bookmark details.

Before using authenticated commands

Check login state before saving, listing, getting, archiving, tagging, or otherwise accessing the user's Slax Reader account:

bash

reader-cli whoami --json

If not logged in, prefer asking the user to run interactive reader-cli login when they can use a terminal. If they cannot complete the interactive flow, offer reader-cli login --api-key <key> --json and ask for an API key only if they explicitly choose that path.

Do not require a login check before login, logout, upgrade --check, skill --check, or skill --sync.

Commands

Task	Command
Log in interactively	`reader-cli login`
Log in non-interactively	`reader-cli login --api-key <key> --json`
Log out	`reader-cli logout --json`
Show account	`reader-cli whoami --json`
Save URL	`reader-cli add <url> --json`
Save with title	`reader-cli add <url> --title "Title" --json`
Save with description	`reader-cli add <url> --description "Description" --json`
Save with tags	`reader-cli add <url> --tags "tag1,tag2" --json`
Archive URL	`reader-cli add <url> --archive --json`
List bookmarks	`reader-cli list --json`
List bookmarks by page	`reader-cli list --page <number> --size <number> --json`
Filter bookmarks	`reader-cli list --filter <all\	inbox\	archive\	starred> --json`
Get bookmark detail	`reader-cli get <bookmark-id>`
Get bookmark as Markdown	`reader-cli get <bookmark-id> --markdown`
Archive a bookmark	`reader-cli archive <bookmark-id> --json`
Unarchive a bookmark	`reader-cli unarchive <bookmark-id> --json`
Star a bookmark	`reader-cli star <bookmark-id> --json`
Unstar a bookmark	`reader-cli unstar <bookmark-id> --json`
Delete a bookmark (recoverable)	`reader-cli delete <bookmark-id> --json`
Restore a deleted bookmark	`reader-cli restore <bookmark-id> --json`
Check CLI update	`reader-cli upgrade --check --json`
Sync skill	`reader-cli skill --sync --json`
Check skill sync	`reader-cli skill --check --json`

Bookmark rules

Preserve the exact URL the user asked to save.
Add --json when you invoke commands that support it so update hints are visible.
Use comma-separated tags with --tags "tag1,tag2".
Use --archive only when the user asks to archive/preserve content or when they explicitly approve archive mode.
If the user provides multiple URLs, save them one at a time so failures are isolated.

Listing and getting bookmarks

Use reader-cli list when the user wants to see their saved bookmarks.
Use reader-cli list --filter inbox, reader-cli list --filter archive, or reader-cli list --filter starred when the user asks for a specific subset.
Use --page and --size when the user asks for pagination or a specific number of results.
Use reader-cli get <bookmark-id> when the user asks to open, inspect, summarize, or read details for a specific bookmark ID.
Use reader-cli get <bookmark-id> --markdown when the user wants the content in Markdown format (e.g. for further processing or display).
Do not guess bookmark IDs; ask the user for an ID or list bookmarks first.

Organizing bookmarks

add <url> --archive is a save-time mode for a new URL; archive <bookmark-id> moves an already-saved bookmark out of the inbox. They are unrelated — do not confuse them.
Use reader-cli archive <bookmark-id> when the user wants to archive a bookmark; use reader-cli unarchive <bookmark-id> to move it back to the inbox.
Use reader-cli star <bookmark-id> / reader-cli unstar <bookmark-id> when the user wants to star or unstar a bookmark.
Use reader-cli delete <bookmark-id> to delete a bookmark — this is recoverable, not permanent. Use reader-cli restore <bookmark-id> to undo it.
Confirm with the user before deleting a bookmark unless they've already clearly asked you to delete it.
Do not guess bookmark IDs for any of these actions; ask the user for an ID or list bookmarks first.

Update and skill hints

reader-cli --json commands may include a top-level _hints object.

`_hints.upgrade`

When _hints.upgrade appears, finish the user's current request first. Then tell the user there is a newer CLI version and offer to run:

bash

reader-cli upgrade

reader-cli upgrade also syncs the AI Agent skill when needed.

`_hints.skill`

When _hints.skill appears, finish the user's current request first. Then tell the user the installed AI Agent skill is out of sync with the CLI and offer the same repair command:

bash

reader-cli upgrade

After the skill is synced, tell the user to restart the Agent so the updated skill is loaded. Resuming an old conversation after restart is fine.

Use reader-cli skill --sync --json only as a direct manual fallback when reader-cli upgrade cannot be used.

Do not silently ignore _hints.

Security rules

Never print the full API key in conversation or command output.
Confirm with the user before saving URLs unless the user has already clearly asked you to save them.
Do not invent URLs. Save only URLs provided by the user or found during an authorized research task.