← 返回 Skills 市场
Milvus
作者
Lumina Wang
· GitHub ↗
· v1.0.0
· MIT-0
230
总下载
0
收藏
1
当前安装
1
版本数
在 OpenClaw 中安装
/install milvus
功能描述
Operate Milvus vector database with pymilvus — collections, vector search, hybrid search, indexes, RBAC, partitions, and more via Python code.
安全使用建议
This skill appears coherent for operating Milvus via Python. Before installing: (1) confirm the install source for pymilvus (prefer the official PyPI/package repository and a known version); (2) be prepared to provide MILVUS_URI (it may include host, port, and credentials) and/or tokens — treat those as secrets; (3) if connecting to a remote Milvus, ensure network access and that you use least-privilege credentials (narrow-scoped API key or limited user) and TLS; (4) if you only need local/dev work, prefer Milvus Lite to avoid sending data to remote endpoints. If you want higher assurance, request the skill's exact installer details (what 'uv' means) and the exact pymilvus version it will install.
功能分析
Type: OpenClaw Skill
Name: milvus
Version: 1.0.0
The milvus skill bundle is a legitimate and well-documented tool for interacting with Milvus vector databases using the official pymilvus library. It provides standard functionality for collection management, vector operations, and RBAC (Role-Based Access Control) as described in SKILL.md. The instructions include safety guidelines, such as requiring user confirmation before destructive operations, and there are no indicators of data exfiltration, malicious execution, or prompt injection.
能力评估
Purpose & Capability
Name/description match the declared requirements: python3 and the pymilvus package are appropriate for operating Milvus. Declared primary credential MILVUS_URI fits the purpose (connection string/endpoint).
Instruction Scope
SKILL.md contains only Milvus-related code examples and instructions (connect, schema, collections, search, RBAC). It does not instruct reading unrelated system files or environment variables beyond connection/auth details.
Install Mechanism
Install spec installs the pymilvus package (expected). The installer kind is listed as 'uv' which is uncommon/ambiguous in the metadata — likely a package install from Python package registry, which is reasonable, but you should confirm the install source is the official PyPI release for pymilvus before installing.
Credentials
Only a single primaryEnv (MILVUS_URI) is declared — appropriate for a database client. The skill references token/user/password as connection parameters in examples but does not require unrelated secrets. Note: connection URIs or token parameters may contain sensitive credentials, so treat MILVUS_URI as secret.
Persistence & Privilege
always is false and the skill is user-invocable; it does not request persistent elevated privileges or write to other skills' configs. No indicators of privileged or forced installation.
如何使用
- 确保已安装 OpenClaw（本地或 Docker 部署）
- 在对话框中输入安装命令：
/install milvus - 安装完成后，直接呼叫该 Skill 的名称或使用
/milvus触发 - 根据 Skill 的参数说明提供必要输入，即可获得结构化输出
版本历史
v1.0.0
Initial release of the Milvus skill — operate Milvus vector databases via Python and pymilvus.
- Connect to Milvus (Lite, Standalone, Cluster, or Zilliz Cloud) with flexible authentication.
- Create collections with auto or custom schemas, including a full range of data types.
- Perform vector insert, upsert, search, query, get, and delete operations.
- Manage indexes, partitions, databases, and control access with RBAC.
- Includes guides for connection, collection management, operation parameters, and code examples.
元数据
常见问题
Milvus 是什么？
Operate Milvus vector database with pymilvus — collections, vector search, hybrid search, indexes, RBAC, partitions, and more via Python code. 它是一个面向 Claude Code / OpenClaw 的 AI Agent Skill 插件，目前累计下载 230 次。
如何安装 Milvus？
在 OpenClaw 或 Claude Code 对话框中运行命令「/install milvus」即可一键安装，无需额外配置。
Milvus 是免费的吗？
是的，Milvus 完全免费，采用 MIT-0 许可证，可自由下载、安装和使用。
Milvus 支持哪些平台？
Milvus 跨平台运行，可在任意部署了 OpenClaw / Claude Code 的环境中使用（cross-platform）。
谁开发了 Milvus？
由 Lumina Wang（@lumina2025）开发并维护，当前版本 v1.0.0。
推荐 Skills