# Grandpa's AI [Health: Active]

**Category:** 🔬 Research  
**Repository:** https://github.com/wescorp/grandpasai-mcp  
**GitHub Stars:** 0  
**Views:** 0  
**Installs:** 0  
**Upvotes:** 0  
**Directory Page:** https://allmcps.com/mcp/grandpa-s-ai

## Description
Search Grandpa's AI research on age reversal, robotics, AI, and space — 5 read-only MCP tools.

## Tools
Capabilities this server exposes over MCP:

- **search_research** — Full-text search across the corpus, ranked by relevance
- **get_article** — Retrieve a full article by path (markdown or summary)
- **list_categories** — All categories & subcategories with article counts
- **get_latest_research** — Most recently updated articles (optionally by category)
- **get_topic_tree** — Full hierarchical topic tree with page URLs

## Claude Desktop Quick Installation
No confirmed install details yet, so no config is published. Follow the setup steps in the repository above.

## Documentation

## What Grandpa's AI does

Grandpa's AI MCP server exposes the research published on grandpasai.com to MCP-compatible agents. The corpus contains more than 230 curated articles covering age reversal, artificial intelligence, humanoid robotics, space colonization, and custom categories. Access is read-only, so the server is suitable for research and information retrieval workflows where an agent needs to inspect source material without making changes.

The Grandpa's AI corpus can be searched by topic or keyword, then explored through categories, article paths, update dates, and a hierarchical topic tree. Article retrieval supports either the full Markdown content or a summary. Topic results include page URLs, which can help an agent connect retrieved material with the corresponding website pages.

## How it works

Remote clients connect to `https://grandpasai.com/mcp` using Streamable HTTP. The endpoint is public and does not require an API key or other authentication. Local clients can instead launch the npm package through stdio, allowing an MCP host to start the server as a local process.

The server provides five MCP tools. Search results are ranked by relevance across the corpus. Other tools expose structured navigation and retrieval functions: categories include article counts, latest research can be filtered by category, and the topic tree represents the full hierarchy of available pages.

## Setup and configuration

For a remote connection, register the endpoint URL in the MCP client's server configuration. No package installation or environment variables are needed for this mode. Claude Desktop, Claude Code, Cursor, VS Code, Codex, and Gemini clients are named as supported connection options in the project documentation.

For local stdio access, install the package globally with `npm install -g grandpasai-mcp`, then run `grandpasai-mcp`. An MCP client can also use `npx -y grandpasai-mcp` as its configured command and argument. The repository identifies the package as `grandpasai-mcp` version 1.0.0.

## Tools and capabilities

- `search_research`: Performs full-text search across the research corpus and ranks matches by relevance.
- `get_article`: Retrieves an article by path as either Markdown or a summary.
- `list_categories`: Lists categories and subcategories with article counts.
- `get_latest_research`: Returns recently updated articles, with optional category filtering.
- `get_topic_tree`: Returns the complete hierarchical topic structure and page URLs.

These tools make Grandpa's AI useful for agents that need to locate background research, inspect a specific article, enumerate available subject areas, or identify recently updated material before answering a question.

## Limitations and notes

Grandpa's AI is limited to read operations over the published corpus; it does not provide tools for writing, editing, or managing articles. Its subject coverage is defined by the available categories and articles on grandpasai.com, so it should not be treated as a general web search service. Remote use depends on the public Streamable HTTP endpoint, while local use requires an environment capable of running the npm package. The project is listed in the Research category and is distributed under the MIT license.

_Full upstream README: https://allmcps.com/mcp/grandpa-s-ai/readme_