Summary
Reading a hundred issues without spending the context on JSON.
General-purpose GitHub servers return every field GitHub has; this one returns five, which is what makes a bulk pass over a repository's issue history fit in a single session. It only lists — nothing here writes.
What it is
Two GitHub listing tools that return only id, html_url, title, body and comment bodies — built for reading a lot of issues or PRs in one session.
What you get
- Issues from a repository, updated at or after a timestamp you give
- Pull requests from the same repository, in the same trimmed shape
- Five fields per item: id, html_url, title, body, comment bodies
- Responses a fraction of the size of the full GitHub API equivalent, because every field outside those five is dropped
- Awareness of GitHub's API rate limits built in
Requirements
A GitHub personal access token as GITHUB_TOKEN, scoped repo or public_repo, plus read:org for organisation repositories. Installed globally from npm as @wipiano/github-mcp-lightweight.
Setup effort
One command plus a key — npm install -g @wipiano/github-mcp-lightweight, then supply credentials