maven-central-publishing
Pass
Audited by Gen Agent Trust Hub on Oct 23, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: No security issues detected. The skill and its associated scripts perform legitimate project configuration analysis and provide guidance for Maven Central publishing.
- [DATA_EXPOSURE_AND_EXFILTRATION]: The Python inspection script
scripts/inspect-project.pyis restricted to local file system reads and regex-based configuration scanning with no network capabilities or exfiltration patterns. - [CREDENTIALS_UNSAFE]: The skill explicitly promotes secure secret management through environment variables and secret stores, and provides a utility to detect accidentally inlined credentials in project files.
- [PROMPT_INJECTION]: The instructions implement a defensive posture by requiring point-of-risk confirmation for sensitive actions and prohibiting the agent from accepting legal terms or publishing on the user's behalf without explicit consent.
Audit Metadata