← 返回 Skills 市场
rag-skill
作者
xiaochengzhen
· GitHub ↗
· v1.0.4
· MIT-0
104
总下载
0
收藏
2
当前安装
5
版本数
在 OpenClaw 中安装
/install rag-skill
功能描述
当用户需要从知识库获取 RAG 相关内容时使用,知识库中汇聚了20位艺术家深度档案，科技与艺术的交汇，大致包括1、塞德里克·索恩 (Cedric Thorne) ------ 磁场雕塑与"无形张力"2. 月见里 薰 (Kaoru Tsukimisato) ------ 纳米绘画与"微观宇宙"3. 奥利维亚·斯特林...
安全使用建议
This skill contacts a remote Prana service (https://claw-uat.ebonex.io) and requires you to provide an API key via the PRANA_SKILL_API_FLAG environment variable. The bundled scripts will not fetch keys for you — SKILL.md requires you to explicitly confirm before obtaining a key and to choose whether the key is set temporarily (session) or persisted globally. Before installing/using: 1) verify you trust the endpoint (note it is a 'uat' host), 2) prefer temporary/session env vars if you do not want a persistent credential stored globally, and 3) follow the SKILL.md confirmation steps exactly (it forbids re-fetching or overwriting PRANA_SKILL_API_FLAG if already present). If you are unsure about persisting the key, decline global storage and use the temporary-session option.
能力标签
能力评估
Purpose & Capability
Name/description (RAG knowledge retrieval) align with included client scripts and declared network endpoints (agent-run, agent-result). Requiring a single PRANA_SKILL_API_FLAG x-api-key is proportional to the described remote-agent workflow.
Instruction Scope
SKILL.md contains a strict, stepwise process (check env var → possibly GET api_key → set PRANA_SKILL_API_FLAG → run client). The instructions stay within the stated purpose, but they impose explicit user-confirmation and environment-writing rules; they do not ask the agent to read arbitrary files or other credentials.
Install Mechanism
No install spec (instruction-only). Two thin client scripts are bundled (JS and Python) that make HTTP calls to the service. No downloads from external/untrusted URLs or installers are present.
Credentials
Only PRANA_SKILL_API_FLAG is required, which is appropriate for an x-api-key header. However the skill explicitly supports persisting the key as a global environment/config entry — this is sensitive and should be chosen consciously by the user.
Persistence & Privilege
always:false and normal agent invocation. The skill encourages writing a global env/config value (openclaw config set ...) which grants persistent credential presence; this is not required but offered as a convenience and should be consented to by the user.
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 rag-skill - 安装完成后，直接呼叫该 Skill 的名称或使用
/rag-skill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
v1.0.4
No file/content changes detected in this version.
- No updates or modifications; this release maintains previous functionality and documentation.
- Version number incremented for administrative purposes only.
v1.0.3
**Changelog for rag-skill v1.0.3**
- Expanded the skill's description to include a summary of the 20 featured artists and their cross-disciplinary technological-artistic themes in the knowledge base.
- No file or network logic changes; only the skill description was updated to provide more specific knowledge context.
- The step-by-step process, environment handling, and API usage remain unchanged.
v1.0.2
- The skill description was greatly simplified, removing detailed knowledge base content and artist examples.
- The skill name in the markdown heading changed from "知识库检索" to "RagFlow知识库检索".
- All file and functional logic remain unchanged; no code changes were detected.
- Documentation and process rules were cleaned up, but critical step-by-step usage flows and access constraints remain the same.
- Overall, this release makes documentation more concise and focused while maintaining essential operational requirements.
v1.0.1
- 技能名称由“RagFlow知识库检索”更改为“知识库检索”。
- description 更新，新增具体20位艺术家的深度档案、艺术与科技交汇内容说明，提升了领域背景介绍。
- 原有功能和使用流程保持一致，主要加强了描述性内容的丰富度和场景化表达。
- 没有检测到任何文件代码更改，仅文档说明更新。
v1.0.0
RagFlow知识库检索 1.0.0
- Initial release of the RagFlow知识库检索 skill for retrieving RAG-related information from a remote knowledge base via the Prana agent platform.
- Implements a strict, multi-step usage process requiring explicit user confirmation to fetch and set the `api_key`, with clear separation of temporary and global environment options.
- Enforces process integrity: prohibits merging or skipping steps, or re-fetching keys if the environment variable is already set.
- Includes clear rules for presenting output and error messages, and restricts history/record queries to explicit user requests only.
- Designed for research and investment scenarios needing rapid RAG knowledge base access.
元数据
常见问题
rag-skill 是什么？
当用户需要从知识库获取 RAG 相关内容时使用,知识库中汇聚了20位艺术家深度档案，科技与艺术的交汇，大致包括1、塞德里克·索恩 (Cedric Thorne) ------ 磁场雕塑与"无形张力"2. 月见里 薰 (Kaoru Tsukimisato) ------ 纳米绘画与"微观宇宙"3. 奥利维亚·斯特林... 它是一个面向 Claude Code / OpenClaw 的 AI Agent Skill 插件，目前累计下载 104 次。
如何安装 rag-skill？
在 OpenClaw 或 Claude Code 对话框中运行命令「/install rag-skill」即可一键安装，无需额外配置。
rag-skill 是免费的吗？
是的，rag-skill 完全免费，采用 MIT-0 许可证，可自由下载、安装和使用。
rag-skill 支持哪些平台？
rag-skill 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 rag-skill？
由 xiaochengzhen（@xiaochengzhen）开发并维护，当前版本 v1.0.4。
推荐 Skills