github-mcp-server

The material states there are no required keys/environment variables, but the claimed features include push, releases, issues, PRs, Actions, secrets, and security operations on GitHub, which normally require an existing GitHub login session or token-based authorization. If a high-privilege GitHub credential is used, there is a standard risk of misuse for repository writes, secret management, and account-level actions.

The material lists no remote endpoint host, but the claimed functionality inherently requires communication with GitHub services and may upload repository content, issue/PR text, release artifacts, or local folder contents. There is no clear evidence in the provided material that data is sent to unknown or unrelated third-party endpoints.

The system flags this tool as executes-code, indicating it executes code or spawns processes locally. Combined with its 'create repos from local folders and push' capability, it likely uses local Git or related system capabilities as part of normal operation. This alone does not justify a high-risk rating, but the runtime environment should be constrained.

The description explicitly says it can create repositories from local folders and push them, which means it at least needs read access to user-selected local directories. It also involves reading and writing GitHub resources such as repositories, PRs, issues, Actions, and secrets. The material does not show requests for local system permissions beyond the stated functionality, but users should avoid selecting sensitive directories or mismanaging sensitive repository resources.

Positive factors include being open source, auditable, and MIT-licensed. However, it comes from a third-party registry, has 0 stars, unknown maintenance status, and no README, which weakens transparency and maturity signals. There is no evidence of closed-source behavior or obvious malicious intent, so caution is more appropriate than a high-risk rating.