numbpill3d/mcp-radar
OtherLive in productiona small github pages site that publishes an auto-updating servers.json.
auto-updating directory of mcp servers + servers.json api
numbpill3d.github.io/mcp-radar/
- HTML80.1%
- Python10.2%
- JavaScript6.1%
- CSS3.6%
1 Review
MCP Radar has a clear and useful purpose: provide a simple directory of MCP servers and a JSON file that other projects can consume. The static GitHub Pages approach keeps hosting costs and maintenance needs low. Its daily update workflow is currently running sucessfully, and the submission process is especially thoughtful. Contributors can submit a server through an issue form, after which automation checks that the repository is public, looks for duplicates, gathers current metadata, and opens a pull request for human review. That is a friendly contribution path for people who do not know Git.
The next priority should be making the data trustworthy for API users. The repository should document the servers.json structure, which fields are required, whether fields may disappear, and how breaking changes will be handled. A schema version and JSON Schema file would let consumers validate the data safely. It would also help to publish the last successful update time, source name, and verification status for every listing. A server being public does not prove it is safe, maintained, or connected to its stated author, so the site should clearly explain what its checks do and do not guarantee.
The update and submission scripts currently have no visible automated test suite. Tests for duplicate URLs, renamed repositories, deleted projects, malformed upstream data, rate limits, and partial source failures would protect the directory as it grows. The workflow should keep serving the last valid dataset when an upstream source fails rather than publishing incomplete data. Finally, the repository does not currently declare a license. Adding one is important because both the site code and the collected dataset are meant for reuse. Overall, MCP Radar is a practical small project with a strong submission workflow. A stable data contract, clearer trust labels, tests, and licensing would make it much safer to build on.
