phaser4-game-architecture
Pass
Audited by Gen Agent Trust Hub on Sep 7, 2026
Risk Level: SAFECOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill provides instructions for the agent to manage development and preview services. This includes checking for existing listening processes on specific hosts and ports, and executing smoke tests using Vite and Phaser. These operations are within the scope of a technical architecture and build-baseline skill.
- [INDIRECT_PROMPT_INJECTION]: The skill is designed to ingest external data to perform its architecture analysis and state management.
- Ingestion points: The skill reads "Work Items", "module audits", "dependency profiles", and other project-related artifacts as specified in the SKILL.md.
- Boundary markers: The instructions mandate that all operations must integrate with a workflow control system (
$phaser4-game-workflow-control) and require explicit approval (A4-A6 levels) for operations with external or destructive effects. - Capability inventory: The skill has the capability to initiate shell commands for starting development servers and verifying build integrity via smoke tests.
- Sanitization: The skill relies on an external auditing mechanism and defined state-gate transitions to validate incoming work items before execution.
Audit Metadata