game-deploy
Warn
Audited by Gen Agent Trust Hub on Mar 13, 2026
Risk Level: MEDIUMEXTERNAL_DOWNLOADSREMOTE_CODE_EXECUTIONCOMMAND_EXECUTIONCREDENTIALS_UNSAFEPROMPT_INJECTION
Full Analysis
- [EXTERNAL_DOWNLOADS]: The skill requires the installation of an external skill from an unverified third-party provider (
heredotnow/skill) to perform its default deployment functionality.\n- [REMOTE_CODE_EXECUTION]: Deployment logic involves executing a shell script (publish.sh) sourced from an unverified external skill, which may perform arbitrary operations on the user's system.\n- [CREDENTIALS_UNSAFE]: Instructions direct users to save sensitive API keys in plain text within a local file (~/.herenow/credentials). While the skill suggests restricting file permissions, storing secrets unencrypted on disk increases exposure risk.\n- [COMMAND_EXECUTION]: The skill frequently invokes CLI tools such asnpm,npx, andghto build and deploy code, which allows for the execution of arbitrary commands within the project environment.\n- [PROMPT_INJECTION]: The skill processes user-provided build directories (dist/) and passes contents to deployment scripts without implementing sanitization or explicit boundary markers, creating a surface for indirect prompt injection.\n - Ingestion points: The
dist/build directory identified inSKILL.md.\n - Boundary markers: Absent from the deployment instructions.\n
- Capability inventory: Shell script execution (
publish.sh), package installation (npx), and network requests (curl).\n - Sanitization: No sanitization or validation of the build artifacts is mentioned.
Audit Metadata