@series-inc/rundot-game-sdk@5.28.1
RUN.world SDK
AI Security Review
scanned 1mo ago · by lpm-firewall-aiBare npm install only prints a hint. Remaining risk is an explicit, consent-gated setup command that injects a bundled SDK docs index into AGENTS.md, CLAUDE.md, and GEMINI.md, which is first-party dual-use rather than a hidden install attack.
Decision evidence
public snapshot- Explicit `rundot-sdk-setup` can create AGENTS.md, CLAUDE.md, and GEMINI.md in the consumer project and inject a bundled docs index after consent.
- Setup recursively walks the project (skipping .git and node_modules) and rewrites every AGENTS.md and CLAUDE.md it finds.
- The same opt-in command can write rundot/cli_hooks.json and add rundot/docs to .gitignore.
- postinstall only prints a TTY hint and performs no network or file writes.
- Setup is not an install hook; without a TTY and without -y it writes nothing, and interactive prompts default to no.
- child_process in copy-docs.mjs is a prepack docs bundler, not consumer install.
- Firebase and RevenueCat values in the Vite plugin are public playground client keys, not harvested secrets.
- Network usage in dist targets vendor Cloud Run, Firebase, and CDN hosts for the documented game playground APIs.
- No runtime self-dependency, eval, or install-time package-manager execution.
Source & flagged code
11 flagged · loading sourcePackage defines install-time lifecycle scripts.
package.jsonView on unpkgFindings
Affected versions and remediation
This report applies to @series-inc/rundot-game-sdk@5.28.1.
See all versions of @series-inc/rundot-game-sdk for other recorded verdicts.
- Review the evidence and your use of @series-inc/rundot-game-sdk@5.28.1 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.