MCP.so
MCP.so is a community-driven directory that indexes over 19,800 MCP Servers, enabling users to discover and integrate third-party servers that extend AI capabilities like Claude with external data sources, tools, and APIs. The platform provides searchable categories, tags, and featured listings to help developers find servers for filesystem access, web scraping, databases, and other integrations. Users can browse available servers, test them in the playground, and submit their own to the open marketplace.
MCP.so is a community-driven marketplace that indexes over 19,800 MCP Servers—integrations that extend AI systems like Claude with access to external data sources, tools, and APIs. The platform serves developers, software engineers, and enterprise AI teams who need to discover and implement Model Context Protocol servers that expose resources like databases, files, and API integrations while maintaining security through the MCP's built-in protocol standards. Users can browse servers across searchable categories and tags, test integrations in the playground, and submit their own servers to the open marketplace at no cost. The directory features curated listings including hosted servers, official integrations, and the latest additions, alongside notable servers like Playwright, Firecrawl, PostgreSQL, and GitLab connectors. MCP.so distinguishes itself as the largest collection of MCP Servers available, providing a central hub where the community can learn how MCP works—a secure, open-source protocol developed by Anthropic that enables AI assistants to connect with external systems through a standardized client-server architecture without requiring API key sharing with LLM providers.
The policy collects 'usage analytics and deployment statistics' and 'technical data' about user interactions, but does n
No training opt-out mechanism was identified in the legal documents.
Moderate risk. Suitable for internal data but avoid confidential information.
The policy collects 'usage analytics and deployment statistics' and 'technical data' about user interactions, but does not explicitly state whether this data is used to train AI models. The ambiguity around how 'technical data' and 'usage metrics' are utilized presents moderate risk.
“Technical Data What We Collect: Information about how you interact with our MCP servers and platform. Purpose: To improve our services, ensure security, and enhance user experience.”
Users grant MCP.so a license to 'host and distribute' their content, which is reasonably scoped to platform operations. However, the license language could be clearer about limitations and does not explicitly restrict other uses, presenting moderate risk.
“By publishing, you grant MCP.so a license to host and distribute your content through our platform.”
Data is shared with 'service providers who help us operate our services (under confidentiality agreements)' and 'when required by law.' The vague definition of 'service providers' and lack of specificity about which third parties presents moderate risk.
“We do not sell or share your information with third parties except: When required by law With your explicit consent With service providers who help us operate our services”
The policy does not specify data retention periods or deletion procedures. While users have the right to 'request deletion of your data,' there is no clarity on retention timelines or automatic deletion policies, creating moderate uncertainty.
“Your Rights You have the right to: Access your personal information Request corrections to your data Request deletion of your data”
The policy clearly states users retain rights to their MCP servers: 'You retain your rights to any MCP servers you publish.' This is explicit and protective of user IP ownership.
“You retain your rights to any MCP servers you publish.”
The policy makes no mention of SOC2, GDPR, HIPAA, ISO certifications, or other compliance standards. This absence is risky for enterprise use, particularly given the handling of potentially sensitive server configurations and user data.
“We implement industry-standard security measures to protect your information and MCP server data.”
Track and measure your AI tools with Olakai's observability platform.
Explore Olakai