Read, write, and search local files with configurable allowlists. The reference filesystem server maintained by the MCP project.

Repository management, file operations, issues and pull requests against the GitHub REST API.

Read-only access to Postgres databases, including schema introspection and parameterized queries.

Channel and DM access for Slack workspaces — read messages, post, search, and manage reactions.

Browser automation and scraping with Puppeteer. Navigate, click, screenshot, and extract page content.

Run SQL queries against a local SQLite database. Useful for scratch analysis and structured local data.

Fetch web content as markdown. Supports HTML to markdown conversion and selective content extraction.

Issue tracking with Linear — create, update, and search issues, projects, and cycles.

Read and write Notion pages and databases. Search, create pages, update properties, and append content blocks.

Manage Stripe resources — customers, products, invoices, payment links — directly from your model.

Pull Figma file context — frames, components, styles — into your model so it can reason about designs.

Inspect and manage Kubernetes clusters — list pods, exec, view logs, and apply manifests.

Query and inspect MongoDB collections. Read-only by default; supports aggregations and indexes.

Read and write to Redis from your model — keys, hashes, streams, and pub/sub.

Run analytical SQL against local Parquet, CSV, and Arrow datasets via DuckDB.

Cypher access to Neo4j graph databases — read schemas, run queries, traverse paths.