io.github.onefreeman1337/osf-data-marketplace

The materials explicitly state there are no required keys or environment variables, and there is no indication that users must provide API keys or system credentials. The description mentions x402 USDC pay-per-record, but the provided materials do not show any requirement to expose wallet private keys, signing authority, or other sensitive credentials.

The tool connects to the remote endpoint api.osf-master-server.com; for a data-marketplace MCP, user queries or request parameters would typically be sent to that service. The materials do not disclose a more detailed data flow, privacy notice, or whether any other third parties receive data, so the scope of outbound data warrants attention.

The system checks indicate this tool has executes-code capability, meaning it runs code or processes locally. The materials do not describe the exact system capabilities, sandbox boundaries, or least-privilege design, but local execution alone is a standard MCP property and is not, by itself, enough to classify as high risk.

The materials only state that it accesses public-domain government/scientific data; they do not specify which local files, cache paths, or persistence locations may be read or written, and there is no stated request for data permissions beyond the claimed function. Since execution-capable MCPs may inherently touch input data and the local runtime environment, it should be treated as having limited but noteworthy data access unless proven otherwise.

Positive factors include distribution via an official registry and updates within the last year, which help reduce risk; however, the source is not auditable, no license is declared, and community adoption is effectively zero stars, reducing verifiability and external scrutiny. Overall, there is no explicit malicious red flag, but the supply-chain transparency is limited and warrants caution.