@66rpg/lumimaker@0.1.46
幻光文游制作器:登录、校验并发布到幻光文游社区;validate-template(TS);login / publish 等转发 lumimaker
AI Security Review
scanned 1mo ago · by lpm-firewall-aiSkill installation is a real, user-invoked write into many AI-editor skill directories, so it is a guarded first-party agent capability rather than install-time hijacking. The AGENTS.md reference is only a template file-presence check, and package.json has no install lifecycle hook.
Decision evidence
public snapshot- The skill subcommand is an explicit CLI path: it prompts for an editor platform unless --platform or --all is already set, then forwards those flags to the native binary and prints copied skill names and target paths.
- skill-targets.json lists project skill directories for many editors, including .agents/skills, .cursor/skills, .claude/skills, and .codex/skills.
- package.json exposes only the lumimaker bin and has build, test, prepack, postpack, and prepublishOnly scripts, with no preinstall, install, or postinstall hook.
- The scanner hit in dist/index.js is the string AGENTS.md inside a template required-file check, not a write to an agent control file.
- No credential harvesting, eval, remote payload fetch, or install-time execution appears in the inspected entrypoints.
Source & flagged code
5 flaggedA single source file combines environment access, network access, and code or shell execution; review context before blocking.
dist/index.jsView on unpkg · L1Findings
Affected versions and remediation
This report applies to @66rpg/lumimaker@0.1.46.
See all versions of @66rpg/lumimaker for other recorded verdicts.
- Review the evidence and your use of @66rpg/lumimaker@0.1.46 before allowing it. Restrict the permissions described in this report.
- Choose an independently verified alternative or release. This report does not establish that other versions are safe.
Evidence last updated: .
Think this verdict is wrong? Dispute it on GitHub or read how reports are produced.