The listing alone is nearly useless for writing anything — you end up fetching each README anyway, one call and one round trip at a time. Returning them together means a digest is one tool call, which is the explicit reason the author gives for building it. The proxy support is not incidental either: it exists because reaching GitHub is not uniform everywhere, and the status tool tells you which way the request is going before you debug a timeout.
A fetcher for the GitHub trending page that returns more than the listing. For each repository it collects the name, link, description and the full README, so a summary or a blog post can be written from one response rather than from a chain of follow-up fetches.
- `get_github_trending` returns the day's trending repositories as Markdown, with the README content included for each one
- `repo_limit` sets how many repositories to fetch, defaulting to 5
- `save_to_file` writes the result to a dated Markdown file and defaults to true, adding a numeric suffix rather than overwriting an existing file
- `use_proxy` routes the fetch through a configured proxy, defaulting to false
- `get_proxy_status` reports the current proxy configuration
Python. Install straight from the repository with pip, or clone and install in editable mode, then have the client run `python -m github_trending_mcp`. Proxy settings come from three environment variables — `PROXY_HOST` defaulting to `127.0.0.1`, `PROXY_PORT` defaulting to `10808` and `PROXY_PROTOCOL` defaulting to `socks5` — which can also be set inline in the client's server entry. Because it parses the trending page, a change to that page's structure will require the parsing logic to be updated.
One command — pip install git+https://github.com/RusianHu/GitHubDailyTrendingFetcher-mcp.git