Back to Plugins

Zilliz

Zilliz Cloud management plugin with 14 skills covering cluster lifecycle, collection schema, vector search, index tuning, bulk import, RBAC, backups, and monitoring.

databasemonitoring
By Zilliz
32Updated 1 month agoApache-2.0

Installation

/plugin install zilliz@claude-plugins-official

How to install

  1. Open Claude Code in your terminal
  2. Run the installation command above
  3. The plugin will be enabled automatically
  4. Use the plugin's features in your Claude Code sessions

Zilliz Cloud Plugin for Claude Code & OpenAI Codex

Your Zilliz Cloud console in your AI coding agent. This plugin lets Claude Code or OpenAI Codex operate the zilliz-cli, bringing the full power of Zilliz Cloud — cluster management, collection operations, vector search, RBAC, backups, and more — into your terminal through natural language.

It ships as a dual-format plugin living in plugins/zilliz/: .claude-plugin/plugin.json for Claude Code and .codex-plugin/plugin.json for Codex, sharing one real skills/ directory — no symlinks, no duplication. Both root .claude-plugin/marketplace.json and .agents/plugins/marketplace.json point their plugin source at ./plugins/zilliz.

What It Does

Instead of memorizing CLI commands or switching to the web console, just describe what you want:

  • "Create a serverless cluster in us-east-1 and set up a collection with 768-dimension vectors"
  • "Search for similar items in my product collection with filter age > 20"
  • "Show me the status of all my clusters and collections"
  • "Set up a daily backup policy for my production cluster with 7-day retention"
  • "Create a role with read-only access to the analytics collection"
  • "Import data from S3 into my embeddings collection"
  • "Check the status of my import job"
  • "Show me this month's usage and invoices"
  • "Spin up a Vector Lake in us-east-1 and create an on-demand cluster with a 30-minute idle TTL"
  • "Refresh my external collection from Vector Lake and tell me when the job finishes"

Installation

From Claude Code Marketplace

Search for zilliz in the plugin directory, or add the marketplace manually:

/plugin marketplace add zilliztech/zilliz-plugin

zilliztech/zilliz-plugin is the GitHub repo. Once added, the marketplace is named zilliztech (from marketplace.json) and the plugin is named zilliz. Then install via the Discover tab in /plugin.

Direct Install

/plugin install zilliz@zilliztech

The format is <plugin>@<marketplace>zilliz@zilliztech, not zilliz@zilliztech/zilliz-plugin (that repo path is not the marketplace name).

Upgrading

/plugin marketplace add does not refresh an already-added marketplace, and /plugin install is a no-op if the plugin is already installed. To pull a new version:

/plugin marketplace update zilliztech
/plugin update zilliz@zilliztech

With OpenAI Codex

Codex reads the same repo via .agents/plugins/marketplace.json, which points at the self-contained Codex plugin under plugins/zilliz/ (its real skills/ directory is shared with the Claude Code build). Add the marketplace:

codex plugin marketplace add zilliztech/zilliz-plugin

Then open /plugins in Codex and install zilliz from the marketplace.

Via codex-marketplace (one-liner)

Alternatively, install directly with codex-marketplace:

npx codex-marketplace add zilliztech/zilliz-plugin --plugins

All skills are shared with the Claude Code build; the quickstart and status skills replace the Claude /zilliz:* slash commands under Codex.

Quick Start

After installing the plugin:

  • Claude Code: run /zilliz:quickstart
  • Codex: invoke the quickstart skill (e.g. "set up the Zilliz CLI")

This guides you through installing the CLI, logging in, and connecting to a cluster.

Commands (Claude Code)

CommandDescription
/zilliz:quickstartInstall CLI, authenticate, and set up cluster context
/zilliz:statusShow cluster status, collections, and statistics

Under Codex these are available as the quickstart and status skills below.

Skills

SkillDescription
zilliz:quickstartOne-shot onboarding: install CLI, authenticate, select and set cluster context
zilliz:statusOverview of context, cluster details, databases, and collections with stats
zilliz:setupInstall zilliz-cli, log in, configure credentials, set active cluster context
zilliz:clusterCreate, list, describe, delete, suspend, resume, or modify clusters (including create-vectorlake for Vector Lake instances)
zilliz:on-demand-clusterCreate, list, describe, or delete on-demand (Vector Lake) clusters with auto-suspend TTL
zilliz:databaseCreate, list, describe, or drop databases
zilliz:collectionCreate, list, describe, drop, rename, load, release, manage aliases, and query per-collection metrics
zilliz:external-collectionTrigger, describe, or list refresh jobs for external collections (Vector Lake-backed)
zilliz:indexCreate, list, describe, or drop indexes
zilliz:partitionCreate, list, load, release, or drop partitions
zilliz:vectorSearch, query, insert, upsert, get, or delete vectors
zilliz:importBulk data import from cloud storage; manage import stages
zilliz:backupCreate, list, describe, delete, export, restore backups, or manage backup policies
zilliz:jobTrack async Cloud Job status (backup, restore, migration, import, clone)
zilliz:user-roleManage database users, roles, passwords, and access privileges
zilliz:monitoringCheck cluster status, collection statistics, and load states
zilliz:billingCheck usage, view invoices, or manage payment methods
zilliz:project-regionManage projects and storage volumes
zilliz:privatelinkManage PrivateLink endpoint services, endpoints, and whitelist

Capabilities

AreaWhat You Can Do
ClustersCreate, delete, suspend, resume, modify
Vector LakeCreate Vector Lake instances; provision, list, describe, and delete on-demand clusters with auto-suspend TTL
CollectionsCreate with custom schema, load, release, rename, drop
External CollectionsTrigger / describe / list refresh jobs against Vector Lake-backed sources
VectorsSearch, query, insert, upsert, delete
IndexesCreate, list, describe, drop
DatabasesCreate, list, describe, drop
Users & RolesRBAC setup, privilege management
BackupsCreate, restore, export, policy management
ImportBulk data import from cloud storage; managed import stages
JobsTrack async operations (backup, restore, migration, import, clone)
PartitionsCreate, load, release, manage
MonitoringCluster status, collection stats, load states, cluster and per-collection time-series metrics
BillingUsage, invoices, payment methods
ProjectsProject and region management
PrivateLinkEndpoint services, project endpoints, whitelist management

Requirements

License

Apache License 2.0

View source on GitHub