Git
Model Context ProtocolRead-only access to a repository's history: status, diffs, commit logs, and blame
Read-only access to a repository's history: status, diffs, commit logs, and blame On the other side: official github server covering repositories, issues, pull requests, code search, and actions
Git reads the repository; GitHub runs the project. Pick by whether the question is local or social.
Local history versus the platform. The Git server reads a repository on disk — log, diff, blame — and never touches a network. GitHub's server covers issues, pull requests, code search, and Actions, and requires a token that can act as you.
| Option | Install | Source | Link |
|---|---|---|---|
| Git | npx -y ./ | Official | Repository ↗ |
| GitHub | hosted endpoint | Official | Repository ↗ |
You want an agent to understand why a file looks like it does, offline and read-only.
The work spans issues, reviews, or CI, not just the working tree.
A project going archived, or a maintainer changing the install path, flips this comparison faster than any feature release. Both entries are re-checked weekly, and the date above tells you how much to trust the details.
Gives an agent scoped read, write, and search access to directories you explicitly allow
Runs read-only SQL against a Postgres database and returns the schema alongside results, which is what makes it usable without pasting table definitions into the prompt
Brave's own server, and the one to use now that the older reference implementation sits in the archived repository
Drives a real browser through accessibility snapshots rather than screenshots, so an agent can click, type, and assert without a vision model
A persistent knowledge graph stored in a local JSON file, exposed through entities, relations, and observations
Talks to the Kubernetes API directly instead of shelling out to kubectl, so an agent can list contexts, read pod logs, and apply manifests without parsing command output