unity-claude-code
Pass
Audited by Gen Agent Trust Hub on May 12, 2026
Risk Level: SAFEPROMPT_INJECTIONCOMMAND_EXECUTION
Full Analysis
- [PROMPT_INJECTION]: The skill exhibits an attack surface for indirect prompt injection as it is designed to ingest and act upon data from external Unity repositories.
- Ingestion points: The skill reads Unity project files (Assets, Packages, ProjectSettings) and list contents via the
scripts/unity-preflight.shscript (referenced inSKILL.md). - Boundary markers: The instructions lack explicit delimiters or warnings to ignore instructions that may be embedded within the project's source code, metadata, or documentation.
- Capability inventory: The skill has the ability to execute shell commands (
bash,node) and file system operations (find,sed). - Sanitization: There is no evidence of sanitization or validation of the content read from the repository before it is processed by the agent.
- [COMMAND_EXECUTION]: The skill executes a local shell script (
scripts/unity-preflight.sh) to perform discovery tasks. While the script itself is benign, the skill also encourages the execution of arbitrary validation commands based on the target repository's content (e.g., "repo-specific build/test commands"). This could allow malicious instructions inside an untrusted repository to achieve command execution if the agent follows them during the validation phase.
Audit Metadata