Awesome MCP Servers
MCP Index website: https://www.tensorblock.co/mcp
Hosted API: https://mcp-index.tensorblock.co
Community cleanup queue: Help verify and improve indexed MCP servers
MCP Maintainer Quick Start
If you maintain an MCP server, the index can give your project a public profile, install-config previews, API metadata, and a README badge.
- Add your server in the best
docs/*.mdcategory, or use the Add MCP server issue form. - Include install, transport, auth, supported clients, docs, license, endpoint, and tool details when you have them.
- After merge, the follow-up comment gives you the public profile URL, API profile URL, install-config links, and badge markdown.
- Add the TensorBlock MCP Index badge to your project README so users can jump back to the indexed profile.
- Claim your profile or send metadata fixes through the claim profile and metadata improvement forms. Valid profile claims and structured metadata updates generate draft metadata PRs; once reviewed, merged, and deployed, the public profile shows maintainer claim status plus install, auth, docs, license, tool, and verification metadata when available. Claim status is separate from TensorBlock verification.
Coverage
This repo currently indexes 7,747 unique MCP server links from the category docs. The README stays lightweight while the full directory lives in docs/*.md, data/catalog.json, and the registry MCP server.
Community Cleanup Queue
Want to contribute without adding a new server? Start with the MCP Index Community Cleanup Queue. It links to live GitHub issue queues for good first metadata fixes, broken entries, catalog-health reports, new server submissions, client config requests, and profile claims.
Cleanup work improves the value of the hosted index: fewer duplicates, fresher source links, better install metadata, clearer categories, and more trustworthy profiles.
How to Participate
This repo is a community directory plus a hosted index. Every useful contribution makes the MCP Index easier for people and agents to search, compare, install, and verify.
Choose the path that matches what you want to do.
If you maintain an MCP server:
- Share your public MCP profile from https://www.tensorblock.co/mcp so users can inspect metadata, install configs, source links, and badges without reading the raw markdown.
- Add your server to the best category under Browse by Category, or use the Add MCP server issue form.
- Improve your existing entry with install command, transport, auth requirements, supported clients, docs URL, license, endpoint, and tool details. Use the metadata issue form if you do not want to open a PR directly; when the profile id and structured values are clear, automation drafts a metadata sidecar PR for maintainer review.
- Claim your TensorBlock MCP profile with the claim profile issue form. Claims show community maintainer ownership metadata and do not imply TensorBlock verification.
- Add the TensorBlock MCP Index badge to your project README so users can jump from your repo to the indexed profile.
If you build MCP clients, agents, or developer tools:
- Use the hosted API at https://mcp-index.tensorblock.co to search servers, fetch normalized profiles, list categories, or generate install-config previews.
- Request another install target with the client config issue form. Include the client name, expected config shape, example config, and official docs. Clear requests generate a draft client-config spec PR so maintainers can review the target before implementation.
- Help improve the config generator for Claude Desktop, Cursor, Codex, VS Code, and future clients.
If you want to improve the index itself:
- Fix duplicate, stale, broken, or poorly categorized entries. Use the broken entry issue form when you want maintainers to triage it. Clear dead-link reports can generate direct cleanup PRs, while duplicate, stale, category, safety, or unclear reports generate draft investigation PRs so cleanup work can be reviewed and tracked.
- Add missing metadata that makes search and install generation more accurate.
- Pick a scoped cleanup task from the community cleanup queue.
- Propose verification signals, ranking improvements, or better category rules. The catalog health checker also runs on a schedule to flag duplicate primary links, missing GitHub repos, archived repos, and disabled repos as broken-entry reports.
- Join the TensorBlock Discord to discuss roadmap work before opening a larger PR.
Issue forms are routed automatically. When you submit a server, metadata update, profile claim, client config request, or broken-entry report, the repo adds the right triage labels and posts the next steps so contributors and maintainers can keep the workflow moving. Server submissions with a clear category can generate draft docs PRs, structured metadata updates or profile claims can generate draft metadata PRs, clear client-config requests can generate draft spec PRs, and clear broken-entry reports can generate either direct cleanup PRs or draft investigation PRs depending on the report type.
New server submissions also get intake status labels so the queue is easier to review at scale:
needs-metadatameans required fields, a valid project URL, or category routing still need contributor/maintainer input.duplicatemeans the submitted project URL already appears in the catalog.ready-for-prmeans automation created or updated a draft docs PR for maintainer review.automation-blockedmeans automation generated a branch but GitHub permissions blocked PR creation.
Maintainers can run MCP Add Server Intake Refresh from GitHub Actions to backfill or repair these labels across open server submissions. Run it with dry_run=true first to preview changes, then rerun with dry_run=false to apply them.
New server entries can be simple, but high-quality metadata makes the profile much more useful. The best entries answer:
- What can an agent do with this server?
- How does a user install or connect to it?
- Does it use
stdio,sse, orstreamable-http? - Does it require an API key, OAuth, bearer token, or no auth?
- Which MCP clients does it support?
- Where are the setup docs, source repo, license, and public endpoint?
After a PR lands on main, the deploy workflow rebuilds the catalog and profiles. The hosted API and public profile pages refresh after the Railway deployment succeeds.
The scheduled catalog health check uses the generated catalog to open catalog-health issues for duplicate links and stale or unreachable GitHub repositories. Those issues feed back into the same broken-entry report workflow, so verified dead links can become direct cleanup PRs and ambiguous cleanup work can become reviewable investigation PRs instead of staying as loose maintainer notes.
TensorBlock MCP Index
This repo is both a community directory and an agent-ready index. Humans add MCP servers in markdown category pages; the indexer turns those entries into structured data that agents can search, inspect, and use to draft install configs.
Hosted MCP Index API
TensorBlock provides the MCP Index API as free community infrastructure. We contribute the compute, hosting, data normalization, and ongoing maintenance needed to make this directory usable by agents and applications without requiring every user to clone the repo or parse markdown.
Public website:
https://www.tensorblock.co/mcp
Base URL:
https://mcp-index.tensorblock.co
Useful endpoints:
GET /- discover available endpoints and current catalog size.GET /v1/categories- list categories with entry counts and source docs.GET /v1/servers?query=postgres&limit=5- search servers by name, description, category, or URL.GET /v1/servers?category=Databases&transport=stdio- filter by category, transport, auth type, and result limit.GET /v1/servers/recent?limit=12- list recently added servers using git-derived catalog timestamps.GET /v1/servers/updated?limit=12- list recently updated entries using git-derived catalog timestamps.GET /v1/servers/{id}- fetch the normalized profile for one MCP server.GET /v1/servers/{id}/badge.svg- render a TensorBlock MCP Index badge for project READMEs.GET /v1/servers/{id}/install-config?client=claude-desktop- generate an MCP client config for Claude Desktop, Cursor, Codex, or VS Code.https://tensorblock.co/mcp/servers/{id}- share a public website profile for an indexed server.
What the API supports today:
- Search MCP servers by keyword.
- Filter by category, transport, auth type, and result limit.
- Browse recently added and recently updated servers.
- Browse all categories with entry counts.
- Fetch a normalized server profile by stable server id.
- Generate install-config previews for Claude Desktop, Cursor, Codex, and VS Code.
We want this registry to support more MCP clients and installation formats over time. If you want another client, package manager, transport, auth flow, or metadata field supported, please open an issue or PR with the expected config shape and examples.
We plan to keep investing in this hosted registry: improving metadata quality, expanding install-config coverage, adding verification signals, and keeping the service available for the MCP community. Contributors are welcome to help build the registry by adding servers, improving metadata, reporting bad entries, and proposing better search or verification workflows. When changes land on main, the deploy workflow rebuilds the catalog and profiles before publishing, so newly merged server entries become searchable after the Railway deployment succeeds.
How entries become index data:
- The source of truth is the category markdown under
docs/*.md. - Optional
data/server-metadata/*.jsonsidecars preserve structured install, transport, auth, client, and license metadata without making the markdown entry long. - Each server entry is parsed from a markdown bullet with a link and description.
npm run catalog:buildmerges markdown entries with sidecar metadata and generatesdata/catalog.json.npm run profiles:buildgeneratesdata/profiles/*.jsonfor stable per-server profiles.- The deploy workflow publishes the refreshed catalog to the hosted API after changes land on
main.
For local development:
npm run registry:mcpstarts a local registry MCP server withsearch_servers,get_server_profile, andget_install_configtools.npm run registry:api:devstarts a local HTTP API for search, category browsing, profiles, and install configs. See TensorBlock MCP Index API.
Contributors still submit normal awesome-list entries, but better metadata makes each entry more useful to agents. See the MCP Index Metadata Contribution Guide for examples.
Add or Improve an Entry
To add a new MCP server:
- Pick the best category from Browse by Category.
- Open that category page under
docs/. - Add one markdown bullet using this format:
- [Server Name](https://github.com/owner/repo): Brief description of what the MCP server lets an agent do. Install: `npx your-package`. - Search the repo for your URL or project name to avoid duplicates.
- Open a pull request.
If you are not sure where the server belongs, use the Add MCP server issue form and we can help route it. When the form has enough information, automation can draft a PR with both the markdown entry and a metadata sidecar.
To improve an existing server, edit the same markdown bullet where the server is listed, or use the metadata issue form with the TensorBlock profile id and structured values such as Install:, Transport:, Auth:, Docs URL:, License:, and Tools:. Clear metadata issues can generate a draft sidecar PR automatically.
data/server-metadata/*.json files are source metadata used by the indexer. Generated files such as data/catalog.json and data/profiles/*.json are maintained by the indexer. If you only add a simple server entry, editing the relevant docs/*.md category page is enough for the PR.
For metadata and indexer examples, see the MCP Index Metadata Contribution Guide.
For maintainers validating a metadata/indexer change:
npm run catalog:build
npm run profiles:build
npm test
npm run typecheck
npm run build
Browse by Category
The README is now a lightweight entry point. Browse the full directory in the category pages below, or use data/catalog.json and the registry MCP server for agent-native search.
