The Agent Skills Directory

[SAFE]: No security threats were identified. The skill performs a static analysis of local files to provide architectural feedback and does not access sensitive credentials, perform network exfiltration, or execute arbitrary code.
[PROMPT_INJECTION]: The skill processes untrusted input from user-provided game files, which introduces a surface for indirect prompt injection.
Ingestion points: Files within the game directory specified by $ARGUMENTS, including package.json and source code files.
Boundary markers: None identified; the agent reads files without explicit delimiters or instructions to ignore nested commands.
Capability inventory: The skill is defined as read-only and lacks capabilities for file modification, network access, or system-level changes.
Sanitization: No content sanitization or filtering is performed on the ingested data prior to analysis.

review-game