Knowz
Enables AI agents to connect to hosted Knowz vaults via MCP for knowledge management, supporting operations like search, save, browse, and amend.
README
Knowz
Knowz AI plugin for DeepSeek Harness. Product name: Knowz. Publisher: Knowz AI.
Connects agents to hosted Knowz vaults over MCP. This plugin is knowledge only. For TDD / quality gates, install KnowzCode separately — do not merge the two.
Official discovery is the GitHub topic dsh-plugin. There is no DeepSeek store.
Install
dsh plugin --profile web add github:knowz-io/deepseek-knowz-plugin
Then create a ukz_ key at https://app.knowz.io/settings/api-keys, set KNOWZ_API_KEY in the environment or $DSH_HOME/.credentials.yaml, and restart dsh web.
export KNOWZ_API_KEY=ukz_…
Do not paste API keys in chat. Do not run /knowz register on an existing Knowz account — that creates a new account. /knowz register is not the existing-user login path.
Package id: dsh-knowz.
Auth
The shipped MCP row interpolates process.env.KNOWZ_API_KEY into Authorization: Bearer. Export + restart is the path that attaches the header. $DSH_HOME/.credentials.yaml is the DSH-managed store; it is not automatically materialized into process.env.
OAuth is not in this first slice.
Personal + business tenants: two MCP rows with unique serverName values and two env refs. Same URL https://mcp.knowz.io/mcp is fine. Duplicate serverName fails at load. See MULTI-ACCOUNT.md.
MCP
Official client: @deepseek-ai/dsh-mcp-client. Hosted endpoint: https://mcp.knowz.io/mcp (HTTP JSON-RPC / streamable HTTP).
No secrets, CLIENT_ID, literal bearer token, or x-api-key in this package.
After the key is set, ask the agent to run the knowz-setup skill to map vaults.
Skills
knowz-ask, knowz-save, knowz-search, knowz-browse, knowz-amend, knowz-setup, knowz-status, knowz-flush, knowz-auto
DeepSeek Harness does not ship Cursor-style slash-command files. Ask the agent to run the skill by name.
License
MIT — publisher Knowz AI. Homepage: https://knowz.io
推荐服务器
Baidu Map
百度地图核心API现已全面兼容MCP协议,是国内首家兼容MCP协议的地图服务商。
Playwright MCP Server
一个模型上下文协议服务器,它使大型语言模型能够通过结构化的可访问性快照与网页进行交互,而无需视觉模型或屏幕截图。
Magic Component Platform (MCP)
一个由人工智能驱动的工具,可以从自然语言描述生成现代化的用户界面组件,并与流行的集成开发环境(IDE)集成,从而简化用户界面开发流程。
Audiense Insights MCP Server
通过模型上下文协议启用与 Audiense Insights 账户的交互,从而促进营销洞察和受众数据的提取和分析,包括人口统计信息、行为和影响者互动。
VeyraX
一个单一的 MCP 工具,连接你所有喜爱的工具:Gmail、日历以及其他 40 多个工具。
graphlit-mcp-server
模型上下文协议 (MCP) 服务器实现了 MCP 客户端与 Graphlit 服务之间的集成。 除了网络爬取之外,还可以将任何内容(从 Slack 到 Gmail 再到播客订阅源)导入到 Graphlit 项目中,然后从 MCP 客户端检索相关内容。
Kagi MCP Server
一个 MCP 服务器,集成了 Kagi 搜索功能和 Claude AI,使 Claude 能够在回答需要最新信息的问题时执行实时网络搜索。
e2b-mcp-server
使用 MCP 通过 e2b 运行代码。
Neon MCP Server
用于与 Neon 管理 API 和数据库交互的 MCP 服务器
Exa MCP Server
模型上下文协议(MCP)服务器允许像 Claude 这样的 AI 助手使用 Exa AI 搜索 API 进行网络搜索。这种设置允许 AI 模型以安全和受控的方式获取实时的网络信息。