Independent editorial recommendation directory

Find the MCPs worth using.

MCP100 is an independent, evidence-led editorial index of MCP servers selected to collectively cover important AI workflows and use cases.

Search by task or industry. Explore recommended MCPs and shortlist candidates. Review the evidence before you connect.

💡 What is MCP?Model Context Protocol securely bridges LLMs (Claude, Cursor) to databases, APIs & tools.

Official & first-party MCPs covered

Independent editorial coverage. No affiliation or endorsement. All trademarks belong to their respective holders.

MICROSOFTGOOGLESTRIPEGITHUBSENTRYFIGMACLOUDFLARE

Independent recommendations · No paid placements

MCP100 decision standard

4 evidence dimensions

Evidence before scores. Fit before popularity.

MCP100 is not a popularity ranking. Popularity may trigger editorial review but never guarantees selection. MCP100 does not fill empty seats simply to reach 100.

See how evidence is reviewed →

Task fit

What should it help with?

Match the MCP to a concrete workflow, client, and environment.

Trust evidence

What does the evidence show?

Review identity, permissions, credentials, installation, and known risks.

Maintenance

Is it still cared for?

Keep source reachability and dated repository or package activity visible.

Adoption

Is there real usage?

Use comparable stars and downloads for discovery, never as proof of safety.

Showing 20 of 83 Recommended MCPs

Three ways to explore

Start with the question you can answer

Find by task

What do you want your AI to accomplish?

Why trust MCP100?

MCP100 is an independent editorial directory of AI tools and MCP servers. Payment cannot buy a recommendation. Every recommendation is based on practical evaluation, real-world usefulness, long-term maintenance, and how well tools work together—not popularity alone.

Learn more about MCP100 →