io.github.lorenzo-cambiaghi/lynx

The materials explicitly state no keys or environment variables are required, and there is no indication that users must provide API keys, tokens, or other sensitive credentials, so credential exposure and abuse risk appears low.

The materials declare no remote endpoints and describe it as a “100% local MCP server”; based on the provided facts, there is no evidence of user data being sent to external services.

The system flags it as executes-code, meaning the tool runs local service/code to provide semantic retrieval. This is normal MCP behavior; there is no evidence of suspicious privileges or system actions beyond its stated purpose, but it should still be treated as a local executable component.

Its stated functions—semantic code search, AST chunking, and a code knowledge graph—imply access to local code repositories for reading and indexing/structuring content. The materials do not show access to unrelated data or excessive permissions, but usage should be limited to the necessary project directories.

It comes from an official registry and has an open-source repository with updates within the last year, which are strong positive signals that lower risk. However, the missing README, undeclared license, and very low community adoption (0 stars) limit auditability and maturity, so caution is still appropriate rather than a high-risk rating.