PostgreSQL
Model Context ProtocolRuns 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
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 On the other side: exposes collections, indexes, and aggregation pipelines so an agent can inspect a database instead of guessing its shape
Postgres unless the data is genuinely document-shaped.
Relational versus document, which decides what the agent can safely ask. Postgres rewards a read-only role and known schema; MongoDB's schema flexibility means the agent discovers the shape at query time, which is useful and easy to misread. Both are read-first tools in practice.
| Option | Install | Source | Link |
|---|---|---|---|
| PostgreSQL | npx -y postgresql://localhost/mydb | Official | Repository ↗ |
| MongoDB MCP | npx -y setup | Official | Repository ↗ |
The schema is stable and the questions are joins and aggregates.
The data is document-shaped and the schema varies by collection.
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.
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
Exposes collections, indexes, and aggregation pipelines so an agent can inspect a database instead of guessing its shape
Gives an agent scoped read, write, and search access to directories you explicitly allow
Read-only access to a repository's history: status, diffs, commit logs, and blame
Searches messages, threads, and channel metadata in a Slack workspace, and can post when given a token with write scope
A code editor built around an agent that can read your whole repo.
The default general-purpose assistant, now with agents and deep research.
Still the strongest aesthetic default for generated imagery.