Search and browse a curated catalog of more than 6,500 AI applications through MCP tools, resources, and prompts.
Copy the AI prompt to install this server into Claude Code, Cursor, or another agent β or use 1-click editor setup below.
We haven't yet run this listing's install command through our automated sandbox check. This isn't a red flag β we're steadily working through the catalog.
π‘ Paste the JSON block into your client's configuration file under mcpServers, then restart the application.
Inspect callable tools, capabilities, and parameters exposed to AI agents by HyperStore MCP.
deficlow/HyperStore-MCP MCP server makes the HyperStore AI application directory available to MCP clients. The catalog contains more than 6,500 curated applications. Agents can use it to find tools for a stated task, inspect an individual application, compare alternatives, or browse the directory by category, audience, and use case.
The server is read-only. It does not require credentials, does not modify catalog data, and does not handle personal information according to the project documentation. It is suitable for assistants that need application discovery rather than execution or direct integration with the listed applications.
An MCP client sends tool, resource, or prompt requests to the local or hosted server. The server acts as an asynchronous wrapper around HyperStoreβs public REST API and returns catalog data to the client over MCP. The request path is: LLM client to MCP server to the HyperStore API.
The hosted Streamable HTTP endpoint is available at https://mcp.store.hypergpt.ai/mcp. For local use, the package runs in stdio mode by default. Self-hosted deployments can also use HTTP or SSE transports. Catalog updates are obtained from the same data source used by the HyperStore website.
The simplest local installation uses uvx:
The package can also be installed with pipx, or run through the published Docker image ghcr.io/deficlow/hyperstore-mcp. The hosted endpoint requires no local installation and is documented as requiring no authentication.
For self-hosted HTTP or SSE deployments, MCP_HOST, MCP_PORT, and LOG_LEVEL can be set. Their documented defaults are 0.0.0.0, 8080, and INFO, respectively. The command-line interface accepts --transport http or --transport sse when a network transport is needed.
The project includes connection examples for Claude Desktop, Claude Code, Cursor, Windsurf, Cline, Zed, Gemini CLI, ChatGPT connectors, the OpenAI Responses API, and the Anthropic Messages API. Client configuration differs depending on whether the local stdio process or hosted URL is used.
deficlow/HyperStore-MCP MCP server exposes 13 tools:
search_apps and ai_search.get_app.list_apps, browse_apps, and category_apps.It also provides three resources: an individual application resource, a category resource, and a full catalog index. Three prompts support guided task discovery, application comparison, and category exploration.
The server only exposes catalog information; it does not run, install, or invoke the applications found in HyperStore. Its operations are read-only and depend on the availability and contents of the public HyperStore API. Local HTTP or SSE hosting requires the operator to configure network binding and port settings as needed.
The project is released under the MIT license. The documentation does not describe API-key authentication, OAuth, paid access, or write operations for deficlow/HyperStore-MCP MCP server.
Factual signals from GitHub, npm, and our automated checks β not a rating.
No reviews yet β be the first to share how this listing worked for you.
Showcase your server listing on GitHub or your project documentation. Embed this dynamic SVG badge to highlight official listing status and live engagement.
[](https://allmcps.com/mcp/deficlow-hyperstore-mcp)<a href="https://allmcps.com/mcp/deficlow-hyperstore-mcp"><img src="https://allmcps.com/api/badge/deficlow-hyperstore-mcp?style=directory" alt="HyperStore MCP on AllMCPs" /></a>